PROJECT_TRACKER.md

PROJECT_TRACKER.md

Kanban board for wepppy work packages and vision items

Last Updated: 2026-04-30 Active Packages: 4 Quick Links: Work Packages Directory | God-Tier Prompting Strategy

Purpose

This tracker provides a high-level view of active and planned work packages for both human and AI agents. When starting a session, agents should check this tracker to understand current initiatives and avoid duplicate work.

How to Use This Tracker

For Agents:

• Check "In Progress" when starting work to see what's active
• Review "Backlog" to understand planned work
• Update package status when starting or completing work
• Add new packages to "Backlog" when scoping new initiatives
• Move packages between columns as they progress

For Humans:

• Get a quick snapshot of active development
• Understand what's coming next
• See what's been recently completed
• Identify opportunities to contribute

Package Lifecycle:

1. Start in Backlog (scoped but not started)
2. Move to In Progress when work begins
3. Move to Done when complete (leave for 30 days, then archive to History)

Kanban Principles Applied

1. Visualize Work

This tracker makes all work visible at a glance, helping agents coordinate and avoiding duplicate effort.

2. Limit Work in Progress

Target: 2-4 active packages maximum to maintain focus and ensure packages complete rather than stall.

Current WIP: 4 packages ✅ Within target range

3. Manage Flow

Monitor how long packages spend in each column:

• Backlog → In Progress: Should happen when capacity available and dependencies met
• In Progress → Done: Target 2-4 weeks for most packages; >6 weeks suggests scope issues

Blockers should be explicitly noted in package tracker.md and surfaced here if they affect multiple packages.

4. Make Process Policies Explicit

All policies are documented:

• Package structure: docs/work-packages/README.md
• Agent workflow: AGENTS.md (Creating a Work Package section)
• Prompt crafting: docs/god-tier-prompting-strategy.md
• Templates: docs/prompt_templates/

When agents encounter unclear policies, they should update the relevant documentation immediately.

5. Implement Feedback Loops

Feedback mechanisms:

• Package tracker.md: Decision logs and progress notes capture what's working/not working
• Lessons learned: Package closure notes document insights for future work
• Agent collaboration: Agents encouraged to suggest improvements when they stumble
• This tracker: Review periodically to identify patterns in blocked/stalled work

6. Improve Collaboratively, Evolve Experimentally

Continuous improvement culture:

• Agents have authority to correct/improve documentation when gaps are found
• Stumbling is a system failure, not agent failure
• Every agent should leave the system slightly better than they found it
• Experimental approaches are encouraged; document outcomes in tracker.md

Feedback welcome: If this tracker format doesn't work, propose improvements in a package tracker or directly update this file.

---

📋 Backlog

Work packages that are scoped but not yet started. Dependencies and prerequisites should be noted.

totalwatsed3 Storage and Optional Terms Contract Hardening

Proposed: 2026-04-29 Size: Medium-High (2-4 focused sessions) Priority: High Link: docs/work-packages/20260429_totalwatsed3_storage_optional_terms/ Description: Define and implement additive optional storage/capacity and runoff-partition terms across WEPP-forest outputs and WEPPpy interchange/totalwatsed3 so water-balance closure interpretation is explicit and backward-compatible.

Scope:

• Finalize optional-term contract (names, units, formulas, null semantics) in totalwatsed spec/docs.
• Harden WEPPpy hillslope parsers for legacy + enriched output layouts.
• Wire optional terms into totalwatsed3 and closure-audit tooling with regression coverage.
• Regenerate and re-audit uncapped-spectacular artifact after rollout.

Strategic Value:

• Reduces repeated analyst/operator confusion around storage semantics (TSW vs Total-Soil Water).
• Preserves legacy compatibility while enabling richer closure diagnostics.
• Creates a durable, documented cross-repo contract for future WEPP output evolution.

Dependencies: Production validation was completed on wepp1 because the run directory is not mounted in the local workspace.

Next Steps: Create a follow-up investigation package if the large H2637/H2809 closure residuals need root-cause analysis beyond this schema/contract work.

---

Deprecate and Remove TauDEM Backend

Proposed: 2025-10-27
Size: Medium (3-5 days)
Priority: Medium
Description: The TauDEM watershed delineation backend is deprecated and should be removed from the codebase to reduce complexity and maintenance overhead.

Scope:

• Remove all code paths related to DelineationBackend.TauDEM in wepppy/nodb/core/watershed.py.
• Delete any TauDEM-specific scripts, configuration, or workflow files.
• Ensure the WBT (WhiteboxTools) backend is the default and fully functional replacement for all use cases.

Strategic Value:

• Reduces technical debt and code complexity.
• Simplifies the watershed delineation logic and configuration.
• Lowers the maintenance burden for both developers and agents.
• Focuses testing and development efforts on the modern WBT backend.

Dependencies: Confirmation that the WBT backend fully covers all necessary functionality previously provided by TauDEM.

Next Steps: Create a work package to analyze the full impact of removal, verify WBT feature parity, and execute the removal.

---

Jinja Template Lint Error Resolution

Proposed: 2025-10-27
Size: Small (1-2 days)
Priority: Low (cosmetic, doesn't affect functionality)
Description: Resolve TypeScript/JavaScript linter false positives when parsing Jinja template syntax ({{ ... | tojson }}) inside <script> tags in .htm templates.

Scope:

• Move dynamic values from inline Jinja expressions to HTML data attributes
• Refactor JavaScript to read from data attributes instead of Jinja-injected constants
• Eliminate lint errors while maintaining template functionality
• Pattern applies to: _base_report.htm, potentially other report templates

Strategic Value:

• Cleaner CI lint output
• Better developer experience (no false positive noise)
• More maintainable separation of template data and JavaScript logic
• Standard pattern for future template development

Dependencies: None

Next Steps: Create work package when bandwidth available; not blocking any current work

---

Rename markdown-extract Repo to markdown-toolkit

Proposed: 2025-10-28
Size: Small (1 hour)
Priority: Low
Description: Rename the rogerlew/markdown-extract repository to rogerlew/markdown-toolkit to reflect that it now includes three tools: markdown-extract, markdown-edit, and markdown-doc.

Scope:

• Rename GitHub repository via Settings → Rename
• Update README.md to reflect new name
• Update any documentation/references in wepppy that point to the old repo name
• GitHub automatically redirects old URLs, so existing links remain functional
• Update tools/README.markdown-tools.md references if needed

Strategic Value:

• Accurate branding reflects toolkit nature (not just extraction)
• Clearer communication to users about available tools
• Better positioning for future tools (e.g., markdown-validate, markdown-toc)
• GitHub redirects preserve all existing links

Dependencies: None (safe operation, GitHub handles redirects)

Next Steps: Quick rename when convenient; very low risk

Kubernetes Migration (Pending)

When resuming Kubernetes work:

• Duplicate static build stage for proxy image
• Use init containers for shared assets
• Eliminate shared volume mounts
• Configure Redis keyspace notifications in ConfigMap
• Set resource limits based on profiling

Health Checks:

• Endpoint: /health
• Returns 200 OK when ready
• Checks Redis connectivity
• Use for liveness/readiness probes

Logging in Production:

• Structured logs to stdout (captured by Docker/K8s)
• Per-run logs in working directory
• Centralized aggregation via Loki/ELK if needed
• Redis status messages ephemeral (72-hour retention)

---

🚧 In Progress

Currently active work packages. Limit to 2-4 packages to maintain focus.

Current WIP Count: 4 packages

---

Hillslope MOFE Daily Closure Audit + Contract Definition

Started: 2026-04-30
Status: Milestones 1-3 complete; full-physics closure rework implemented and validated in tool tests
Size: Medium-High (2-4 focused sessions)
Owner: Codex
Link: docs/work-packages/20260430_hillslope_mofe_daily_closure_audit/
Description: Define a source-backed MOFE water-balance contract from /workdir/wepp-forest and implement a dedicated hillslope_mofe_daily_closure_audit tool with tests and run evidence, now explicitly reporting full-physics exported-term closure residuals plus implied unresolved terms.

Current Status:

• Contract milestone and required subagent gate are complete (artifacts/20260430_contract_review_disposition.md).
• Tool implementation now includes full-physics closure diagnostics (RM + UpStrmQ + SubRIn inputs; QOFE + latqcc + Dp + ET + Tile + ΔStorage outputs) and MOFE chain transfer checks.
• Regression test suite for the new tool passes (tests/tools/test_hillslope_mofe_daily_closure_audit.py).
• Remaining lifecycle work: evaluation artifact refresh and independent implementation-review disposition artifact.

Next Steps:

1. Re-run required closure-audit regression guard suites.
2. Refresh evaluation summary/artifacts under package artifacts/.
3. Complete and record implementation review disposition artifact.

---

Watershed Centroid Persistence Hardening for Climate Build Reliability

Started: 2026-04-22
Status: Implementation complete with targeted validation; global suite closure blocked by unrelated Geneva failure
Size: Medium-High (2-5 focused sessions)
Owner: Codex
Link: docs/work-packages/20260422_watershed_centroid_persistence_hardening/
Description: Harden watershed centroid durability and climate-call-site contracts to prevent build_climate_rq failures when watershed.nodb persists with missing centroid state despite successful abstraction artifacts.

Current Status:

• Watershed centroid contract now supports repair-or-fail semantics with typed WatershedCentroidStateError.
• Climate/station centroid consumers now use require_centroid() instead of nullable unpack paths.
• NoDb stale-write rejection is active and covered by regression (NoDbStaleWriteError).
• abstract_watershed_rq now verifies persisted centroid durability, performs one bounded repair attempt, and fails typed if durability still fails.
• Targeted validation passed (43 passed) across nodb/rq/climate touched modules.
• Full-suite gate is currently blocked by unrelated existing failure in tests/nodb/mods/geneva/test_geneva_wp09_end_to_end.py::test_wp09_watershed_warning_thresholds_propagate_to_results_query_report[...] (KeyError: 'severity').

Next Steps:

1. Fix unrelated Geneva worktree failure (KeyError: 'severity' in test_geneva_wp09_end_to_end.py).
2. Re-run wctl run-pytest tests --maxfail=1 and close package after global gate clears.

---

Jagged Hyperpigmentation Hillslope Ablation Queue (H3507, H1271)

Started: 2026-04-22
Status: Observe-only + first hypothesis lanes complete; Windows comparator baseline established
Size: Medium (2-3 focused sessions)
Owner: Codex
Link: docs/work-packages/20260422_jagged_hyperpigmentation_hillslope_ablation_queue/
Description: Prepared next ablation campaign slice for run jagged-hyperpigmentation/disturbed9002-10-mofe, focused on anomalous element.dat signatures and sediment concentration in hillslopes H3507 and H1271.

Current Status:

• Incident package initialized at /workdir/wepp-forest/docs/ablation/20260422_jagged-hyperpigmentation_hillslope_elementdat-stars/.
• Source artifacts staged from wepp1:/geodata/wc1/runs/ja/jagged-hyperpigmentation.
• Baseline local replays complete for p1271 and p3507; source-vs-staged signature scans recorded (C099, C100).
• blarhg Windows comparator binary inventory captured (SHA256 07d348d5f9ebff607b6f8e15bea8647410a080c8451b9017f74a9475f39c569d, timestamp 2026-04-18T02:41:59Z) as case C101.
• Observe-only and first hypothesis lanes completed (C101-C106), including Windows target replays and cross-lane signature census.
• Decision recorded: use C:\src\wepppy-win-bootstrap\bin\wepppy-win-bootstrap.exe as parity comparator baseline for this incident.

Next Steps:

1. Execute the next one-change hypothesis lane group using the pinned Windows comparator baseline.
2. Isolate minimal causal delta for source-only starred signatures.
3. Publish keep/rollback recommendation for next implementation package.

---

markdown-doc Toolkit Integration

Started: 2025-10-25
Status: Phase 3 Complete — Integration Active (Phase 4 pending telemetry + RFC decisions)
Size: Large (Phases 1-3: 6 days; Phase 4: TBD)
Owner: gpt-5-codex (Phase 1-3 implementation), GitHub Copilot (Claude 4.5 Sonnet - spec, testing, integration)
Link: docs/work-packages/20251025_markdown_doc_toolkit/
Description: Comprehensive Rust CLI for documentation management with catalog generation, link validation, safe file moves, TOC maintenance, and reference finding. Phase 4 (search & indexing) pending go/no-go decision.

Completed Deliverables (Phases 1-3):

• ✅ catalog - Generate DOC_CATALOG.md with file list + TOCs
• ✅ lint - Validate links (broken-links, anchors, hierarchy, required-sections modes)
• ✅ toc - Generate/update table of contents with multiple styles
• ✅ mv - Move/rename files with automatic link updates across workspace
• ✅ refs - Find references to files/sections (file graph + forward/reverse links)
• ✅ validate - Config-driven checks with severity tuning
• ✅ wctl integration - 6 doc-* commands: doc-lint, doc-catalog, doc-toc, doc-mv, doc-refs, doc-bench
• ✅ CI/CD integration - docs-quality.yml workflow with SARIF upload, Rust checks, telemetry
• ✅ Comprehensive documentation - tools/README.markdown-tools.md, CI/CD strategy updates
• ✅ Telemetry collection active (started 2025-10-31, logs to telemetry/docs-quality.jsonl)

Phase 4 Scope (Pending Go/No-Go Decision 2025-11-18):

• 🔮 search - Full-text search with TF-IDF ranking (<500ms target)
• 🔮 Index builder with persistent caching (<5s rebuild target)
• 🔮 JSON output mode for programmatic consumption

Current Status:

• RFC Decision Gate (Due 2025-11-08): 4 decisions pending (link graph caching, CI bench cadence, release comms, Phase 4 scope validation)
• Telemetry Collection (Due 2025-11-18): Gathering baseline data (≥2 weeks from 2025-10-31) to validate Phase 4 justification
• Integration Finalization (Due 2025-11-08): Onboarding docs, release notes, RFC outcomes

Strategic Value:

• Delivered (Phases 1-3): Reduces doc maintenance by ~70%, prevents broken links in CI, enables safe refactoring, improves link graph visibility
• Potential (Phase 4): Fast semantic search across 388+ docs, programmatic query support, reusable index for tooling integration

Dependencies:

• Phase 4 blocked on: Telemetry data maturity (≥2 weeks), RFC decisions, Phase 4 open questions resolution

Next Steps:

1. Immediate (Nov 2-8): Resolve RFC decisions (4 decisions), finalize integration (onboarding docs, release notes)
2. Phase 4 Gate (Nov 18): Review telemetry data, resolve Phase 4 open questions (index storage, watch mode, UI integration, search patterns), make go/no-go decision
3. If Phase 4 Greenlit: Create Phase 4 work package structure, draft agent prompts, assign ownership, schedule M1 target (~Dec 6)

---

✅ Done

Recently completed work packages. Archived immediately upon completion.

Uncapped-Spectacular H2637 Ablation Campaign (2026-04-30)

Status: ✅ COMPLETE

Link: docs/work-packages/20260430_uncapped_spectacular_h2637_ablation_campaign/

Lifecycle: Backlog -> In Progress -> Done (2026-04-30)

Summary: Completed end-to-end attribution campaign for uncapped-spectacular hillslope H2637, including required Windows comparator execution on blarhg with C:\\src\\wepppy-win-bootstrap\\bin\\wepppy-win-bootstrap.exe. Incident package 20260430_uncapped-spectacular_h2637_hillslope_closure-spike was executed/finalized in /workdir/wepp-forest/docs/ablation/ with lane matrix C000 (production binary replay), C010 (historical comparator), and C020 (Windows comparator). Day-44 legacy closure spike was reproduced in source + C000 and not reproduced in comparator lanes; day-45 remained near zero. Package artifacts now include artifacts/evaluation_summary.md and artifacts/incident_snapshot/ with matrix, notes, summaries, and integrity records.

Hillslope Daily Closure Audit Tool (MOFE + Single OFE) (2026-04-30)

Status: ✅ COMPLETE

Link: docs/work-packages/20260430_hillslope_daily_closure_audit/

Lifecycle: Backlog -> In Progress -> Done (2026-04-30)

Summary: Closed with a new repeatable hillslope closure audit CLI (tools/hillslope_daily_closure_audit.py) that operates directly on interchange artifacts (H.wat, H.pass, optional H.soil/H.element) for one hillslope selected by --wepp-id or --topaz-id. The tool preserves MOFE accounting by using outlet-OFE-only latqcc and PASS runvol for runoff depth. Regression coverage (tests/tools/test_hillslope_daily_closure_audit.py) validates single-OFE closure arithmetic, MOFE outlet/runoff behavior, selector paths, and invalid selector rejection; existing watershed closure-audit tests remained green. Real-run artifacts were generated for six exemplars (uninsured-deformation H78/H43/H97 and bovine-clipboard H1/H2/H3), plus a topaz-selector verification run, with consolidated stats in artifacts/evaluation_summary.{md,csv}. Independent review findings were dispositioned in artifacts/20260430_code_review_disposition.md.

WEPP Gregorian Leap-Year Contract and Centurial Tolerance (2026-04-30)

Status: ✅ COMPLETE

Link: docs/work-packages/20260430_wepp_gregorian_leap_tolerance/

Lifecycle: Backlog -> In Progress -> Done (2026-04-30)

Summary: Completed leap-year contract correction and compatibility preservation across WEPP source and vendored artifacts. Updated active runtime leap call sites in /workdir/wepp-forest/src (stmget.for, wshpas.for, contin.for, wshdrv.for) to Gregorian classification (/4 except /100 unless /400) while retaining non-400 centurial day-366 tolerance in day-count/loop branches. Rebuilt binaries, validated required wepp-forest gates (smoke for wepp and wepp_hill, hillslope watchlist 12/12, ablation artifact policy, full pytest -q with 79 passed, 2 warnings), and captured centurial controls (year100/365 false warning removed, year100/366 tolerated success, year2000/366 leap success). Released/vendored wepp_260430 and wepp_260430_hill, synced changelog copy into wepppy, and passed post-vendor provenance/smoke/focused regressions (8 passed).

Geneva HRU Peak Runoff and Event Erosion Enablement (2026-04-30)

Status: ✅ COMPLETE

Link: docs/work-packages/20260429_geneva_hru_peak_event_erosion/

Lifecycle: Backlog -> In Progress -> Done (2026-04-30)

Summary: Closed the Geneva HRU-local peak runoff substrate package end to end. Rust geneva_core run_batch now emits per-HRU peak_runoff_m3_s and time_to_peak_minutes using per-HRU incremental excess + HRU area through the existing Geneva unit-hydrograph convolution path (no watershed area-splitting). PyO3 bridge tests were updated, cli_revision_rust release artifacts were rebuilt/synced to both runtime locations, and matching SHA-256 provenance was recorded. WEPPpy now materializes measure_id=hru_peak_runoff (unit=m3_s) in geneva/hru_event_measure_rows.parquet, HRU-map measure validation accepts hru_peak_runoff while still rejecting watershed-only peak_discharge, and Geneva Interactive Summary HRU Choropleth controls now expose HRU peak runoff. Section 12.4/12.5 specification updates, validation summary, code review, QA review, and dedicated security review artifacts are complete. Required gates passed except known unrelated frontend lint baseline in controllers_js/__tests__/landuse_map_inline.test.js, which is documented in the package validation artifact.

NOAA Atlas 14 Retry Backoff Hardening for Climate Artifact Export (2026-04-30)

Status: ✅ COMPLETE

Link: docs/work-packages/20260429_noaa_atlas14_retry_backoff/

Lifecycle: Backlog -> In Progress -> Done (2026-04-30)

Summary: Completed bounded exponential retry/backoff hardening for optional NOAA Atlas 14 climate artifact export. download_noaa_atlas14_intensity now applies parameterized retries (timeout=30s, attempts=3, base=1.0s, cap=8.0s; env-overridable via WEPPPY_NOAA_ATLAS14_*) and logs attempt/backoff/exhaustion context while preserving non-fatal exhaustion behavior. ValueError no-coverage remains immediate non-retryable. Deterministic regressions cover transient failure then success, bounded retry exhaustion, no-coverage immediate return, timeout propagation, cap-hit backoff, and invalid-env fallback. Targeted validation passed: wctl run-pytest tests/nodb/test_climate_artifact_export_service.py --maxfail=1 (12 passed).

Uncapped-Spectacular totalwatsed3 Runoff Reconciliation (2026-04-29)

Status: ✅ COMPLETE

Link: docs/work-packages/20260429_uncapped_spectacular_totalwatsed3_runoff_reconciliation/

Lifecycle: Backlog -> In Progress -> Done (2026-04-29)

Summary: Closed a production runoff-depth reconciliation package for uncapped-spectacular. Corrected totalwatsed3 runoff derivation to use runvol/Area*1000 (instead of QOFE depth basis), added regression coverage, and introduced a repeatable CLI audit (tools/totalwatsed3_daily_closure_audit.py) with unit tests. Performed a no-restart wepp1 runtime hotfix and regenerated /geodata/wc1/runs/un/uncapped-spectacular/wepp/output/interchange/totalwatsed3.parquet; container uptime remained unchanged. Post-refresh verification confirmed max(abs(Runoff - runvol/Area*1000)) = 0.0, and closure-audit outputs were captured under both wepp1 run artifacts and package artifacts.

Geneva Interactive Summary HRU Choropleth Series (2026-04-29)

Status: ✅ COMPLETE

Link: docs/work-packages/20260428_geneva_hru_choropleth_series/

Lifecycle: Backlog -> In Progress -> Done (2026-04-29)

Summary: Closed WP01-WP04 end to end. WP01 finalized contracts and preserved watershed-only peak_discharge; WP02 added HRU event-measure artifact/query support and enforced map-scope validation; WP03 delivered Geneva deck.gl HRU choropleth UI with themed controls and runoff winter palette; WP04 executed validation/docs/release closure. Required scoped pytest, targeted Geneva JS tests, docs lint, and git diff --check passed. Known unrelated wctl run-npm lint baseline in wepppy/weppcloud/controllers_js/__tests__/landuse_map_inline.test.js was reproduced and documented as external to this series.

Geneva Preflight Checklist Freshness Integration (2026-04-29)

Status: ✅ COMPLETE

Link: docs/mini-work-packages/20260429_geneva_preflight_checklist_execplan.md

Lifecycle: Backlog -> In Progress -> Done (2026-04-29)

Summary: Closed the Geneva preflight checklist freshness package end to end. Added dedicated TaskEnum.run_geneva (🐈) ownership, stamped completion on successful Geneva batch runs, and wired Geneva into TOC emoji mapping (#geneva) plus preflight checklist payloads (checklist.geneva). preflight2 now evaluates Geneva freshness against build_landuse, build_soils, build_climate, and conditionally init_sbs_map when attrs:has_sbs == "true". Stale invalidators now clear timestamps:run_geneva at documented boundaries: Geneva config diffs, Geneva CN-table modify/reset, Geneva prepare/panel/workflow enqueue paths, SBS upload/remove/uniform/class edits, and climate/landuse/soils enqueue paths. Follow-up hardening stamped TaskEnum.init_sbs_map on Disturbed/BAER SBS mutation paths that set attrs:has_sbs=true, added Geneva-run legacy backfill for historical has_sbs=true runs missing timestamps:init_sbs_map, and completed a one-time Redis sweep backfilling 99 legacy SBS runs. Authenticated smoke on onshore-xenophobia/disturbed9002_wbt confirmed websocket checklist.geneva=true and Geneva TOC emoji metadata present in run-page bootstrap. Queue dependency artifacts were synchronized (job-dependency-graph.static.json + job-dependencies-catalog.md) and wctl check-rq-graph is green. Targeted validation gates passed; optional full-suite confidence gate still fails on unrelated existing tests/nodb/test_base_boundary_characterization.py::test_dump_forces_monotonic_signature_after_second_same_size_rewrite.

WEPP Interchange Dependency Race Guard (2026-04-29)

Status: ✅ COMPLETE

Link: docs/work-packages/20260428_wepp_interchange_dependency_race_guard/

Lifecycle: Backlog -> In Progress -> Done (2026-04-29)

Summary: Closed the WEPP interchange race-hardening package end to end. Queue wiring now enforces deterministic ordering so _post_watershed_interchange_rq cannot start before _build_hillslope_interchange_rq in enqueue_wepp_pipeline and enqueue_wepp_noprep_pipeline, removing the known H.wat.parquet.tmp commit race window. Regression coverage was expanded in tests/rq/test_wepp_rq_pipeline.py to assert dependency identity across all helpers that enqueue _post_watershed_interchange_rq, including watershed-only paths that remain cleanup-only by design. Queue dependency artifacts (wepppy/rq/job-dependency-graph.static.json, wepppy/rq/job-dependencies-catalog.md) were synchronized and validated (wctl check-rq-graph up to date). Independent reviewer and qa_reviewer artifacts were completed, the one medium QA finding was resolved in-package, and the dedicated security review gate passed with no unresolved medium/high findings.

Geneva Storm Shape Control (2026-04-28)

Status: ✅ COMPLETE

Link: docs/work-packages/20260428_geneva_storm_shape_control/

Lifecycle: Backlog -> In Progress -> Done (2026-04-28)

Summary: Closed the Geneva storm-shape package end to end across /workdir/wepppy and /workdir/wepppyo3. Added closed-enum Storm Shape support (uniform, neh4_type_b, type_i, type_ia, type_ii, type_iii) through UI, controller payloads, Python schemas/services/reports, and Rust kernel hyetograph dispatch. Source gating requirements were satisfied first with checked-in raw WinTR-20 table payload, normalized CSV, and metadata under geneva_core/resources/, including Type II embedded-duration validation within <= 0.003 absolute tolerance. Package closure also addressed reviewer/QA findings: kernel callable availability, panel/run-batch distribution consistency, non-divisible timestep rejection, stale-summary suppression, positive-depth contract alignment, and explicit legacy-uniform warning surfacing. Required validation commands passed except an unrelated pre-existing JS lint issue in controllers_js/__tests__/landuse_map_inline.test.js; reviewer, QA, and validation artifacts are recorded under the package artifacts/ directory.

wepppyo3 Native Substrate Repositioning (2026-04-28)

Status: ✅ COMPLETE

Link: docs/work-packages/20260428_wepppyo3_repositioning/

Lifecycle: Backlog -> In Progress -> Done (2026-04-28)

Summary: Closed the docs-first wepppyo3 repositioning package. /workdir/wepppyo3/README.md now positions the repo as WEPPpy's native kernel and interchange substrate, and new canonical docs define the module registry, architecture boundaries, release provenance, and claim discipline. WEPPpy architecture, package README, root README, and dependency-standard references now point to this posture instead of describing wepppyo3 as a generic accelerator bundle. Validation passed with scoped WEPPpy doc-lint, git diff --check in both repos, uk2us preview for changed Markdown, and manual wepppyo3 relative-link validation. Runtime behavior and release binaries were not changed; the ExecPlan is archived under prompts/completed/.

RQ Scoped Stale NoDb Cache Guard Priority 2 (2026-04-28)

Status: ✅ COMPLETE

Link: docs/work-packages/20260428_rq_scoped_stale_cache_guard_priority2/

Lifecycle: Backlog -> Done (2026-04-28)

Summary: Closed Priority 2 scoped stale-cache guard conformance across deferred RQ module families. Implemented exact pre-hydration guards for WEPP orchestration (wepp.nodb), SWAT build/run/interchange (swat.nodb), Omni scenario/contrast orchestration and contrast deletion (omni.nodb), PATH CE orchestration (path_ce.nodb + omni.nodb), Roads prepare/run (roads.nodb), Geneva prepare/panel/batch (geneva.nodb), and fork undisturbify destination-run mutations (new_runid scopes: ron.nodb, disturbed.nodb, landuse.nodb, soils.nodb). Targeted tests assert scoped pup_relpath values and guard-before-hydration ordering while preserving existing lock/archive/status/timestamp/enqueue/clone/delete/autocommit/runtime-lock behavior. All required focused pytest suites, package doc-lint, and git diff --check passed; ExecPlan archived under prompts/completed/.

RQ Scoped Stale NoDb Cache Guard Follow-Ups (2026-04-28)

Status: ✅ COMPLETE

Link: docs/work-packages/20260428_rq_scoped_stale_cache_guard_followups/

Lifecycle: Backlog -> Done (2026-04-28)

Summary: Closed the scoped RQ stale-cache guard follow-up package. wepppy/rq/project_rq.py now clears exact scoped NoDb cache entries before mutable hydration for Priority 0 project-prep paths and simple Priority 1 mod paths (rangeland_cover, treatments landuse/soils, ash, debris flow, RAP TS, OpenET TS, POLARIS, RUSLE). Regression coverage asserts exact scopes, archive/root rejection before cache clear, guard-before-hydration ordering, and representative status/timestamp/enqueue behavior. Priority 2 WEPP/SWAT/Omni/PATH CE/Roads/Geneva/fork-undisturbify candidates were split/deferred with rationale because they need module-specific orchestration tests. Required targeted pytest, package doc-lint, and git diff --check passed; the ExecPlan is archived under prompts/completed/.

build_soils_rq Stale NoDb Cache Guard (2026-04-28)

Status: ✅ COMPLETE

Link: docs/work-packages/20260428_build_soils_rq_stale_cache_guard/

Lifecycle: Backlog -> Done (2026-04-28)

Summary: Closed the scoped soils RQ stale-cache guard package. build_soils_rq now clears only soils.nodb cache inside the existing soils directory-root lock callback and immediately before Soils.getInstance(wd).build(). Targeted regression coverage asserts cache-clear ordering, scoped key usage, unchanged archive-root rejection before cache clear/hydration, and unchanged success-path status/timestamp order. Required targeted pytest suites, package doc-lint, and git diff --check passed; the ExecPlan is archived under prompts/completed/.

RQ WEPP Subwta Precondition Contract Enforcement (2026-04-27)

Status: ✅ COMPLETE

Link: docs/work-packages/20260427_rq_subwta_precondition_contract/

Lifecycle: Backlog -> In Progress -> Done (2026-04-27)

Summary: Closed strict RQ subwta precondition enforcement for run-wepp and run-wepp-watershed. Missing watershed.subwta now returns canonical HTTP 409 with error.code="invalid_watershed_abstraction_state" before payload mutation, batch/base acknowledgement, or enqueue. Regression coverage locks normal, batch, _base, and checkbox_wepp_watershed=false behavior for both run endpoints, while prep-wepp-watershed keeps its existing missing-subwta prep path. Canonical response docs and schema-default error metadata now include the strict contract and recovery notes; required reviewer/QA findings were dispositioned before closure. The ExecPlan is archived under prompts/completed/.

WEPP Runner Traceability Hardening (Hillslope + Watershed) (2026-04-27)

Status: ✅ COMPLETE

Link: docs/work-packages/20260427_wepp_runner_traceability_hardening/

Lifecycle: Backlog -> In Progress -> Done (2026-04-27)

Summary: Closed the single-rollout traceability package for continuous run_hillslope and run_watershed only. run_watershed now emits startup context parity (runs_dir, run_file, err_file, cmd, attempt=1/1), both in-scope methods log cached binary identity fields (binary_path, SHA256, size, mtime, status/error fallback), and both start a Linux best-effort bounded D-state watchdog controlled by documented env vars. Watershed close-path I/O failures now emit classified close_path_failure diagnostics, including classification=stale_file_handle for the production NFS signature. Targeted tests cover startup parity, binary identity/fallback, watchdog emit/no-emit behavior, and close diagnostics. Post-closure operator addendum executed in-package: built/vendored wepp_260427 + wepp_260427_hill (provenance recorded as 6bb872ca (dirty tree)), preserved -g/-fbacktrace makefile traceability defaults, synced vendored changelog, and documented dirty-cycle guidance for pmxelm.inc/pmxhil.inc/pmxpln.inc/pntype.inc. The ExecPlan is archived under prompts/completed/.

Peridot vs WEPPpy Python Abstraction Benchmark (2026-04-27)

Status: ✅ COMPLETE

Link: docs/work-packages/20260426_peridot_python_abstraction_benchmark/

Lifecycle: Backlog -> In Progress -> Done (2026-04-27)

Summary: Closed as a valid comparator-failure benchmark attempt, then updated with a post-close rough benchmark addendum. The package rediscovered the legacy Python comparator as WatershedAbstraction(topaz_wd, wat_dir) and the NoDb wrapper path _topaz_abstract_watershed(), selected the in-repo wepppy/_tests/feverish-lamp TOPAZ fixture, and copied inputs into isolated scratch directories. Initial Python execution failed with a NumPy casting error in wepppy/topo/watershed_abstraction/support.py::cummnorm_distance(). The post-close remediation fixed that failure plus legacy channel GeoJSON serialization, then collected rough smoke numbers with exact parity explicitly out of scope: Python mean 2.368s, Peridot mean 0.162s, about 14.6x faster for Peridot on this tiny fixture and command path. The ExecPlan is archived under prompts/completed/. Follow-up work remains scoped to fixture curation, parity cleanup, and binary provenance before publication-grade benchmark claims.

Peridot Runtime Contract Hardening (2026-04-26)

Status: ✅ COMPLETE

Link: docs/work-packages/20260426_peridot_runtime_contract_hardening/

Lifecycle: Backlog -> Done (2026-04-26)

Summary: Closed cross-repo runtime/schema hardening for Peridot and WEPPpy. Peridot CLI entrypoints now propagate underlying abstraction io::Result<()> errors so abstract_watershed and wbt_abstract_watershed return non-zero on propagated write-stage failures. Peridot field_flowpaths.csv now uses unique headers with parent topaz_id and flowpath-record flowpath_topaz_id. WEPPpy post-processing normalizes new flowpath_topaz_id and historical pandas-mangled topaz_id.1 schemas to canonical field_flowpaths.parquet while rejecting ambiguous mixed inputs. Canonical Peridot/WEPPpy docs and package validation artifacts were updated. Follow-up Peridot commit e09f54c closed the initially observed support/raster full-suite failures, and local cargo test is now clean.

Peridot Documentation Repositioning and Adoption Visibility (2026-04-26)

Status: ✅ COMPLETE

Link: docs/work-packages/20260426_peridot_documentation_repositioning/

Lifecycle: Backlog -> In Progress -> Done (2026-04-26)

Summary: Closed cross-repo documentation repositioning for Peridot. /home/workdir/peridot/README.md now frames Peridot as an explicit graph abstraction shift rather than a simple TOPAZ/TOP2WEPP modernization, with replacement boundaries, canonical docs, and communication kit. New Peridot docs cover the watershed output contract, benchmark/claim discipline, prepwepp/TOPAZ migration, and operations validation. WEPPpy references now point to canonical Peridot docs and avoid unqualified speedup claims. Package artifacts record claim provenance, validation, and follow-up runtime/schema gaps.

NoDb Atomicity + Observability Follow-Ups (RQ Engine) (2026-04-26)

Status: ✅ COMPLETE
Link: docs/work-packages/20260425_nodb_atomicity_observability_followups_a/
Lifecycle: Backlog -> In Progress -> Done (2026-04-26)
Summary: Closed end-to-end across six milestones. Delivered scoped cross-controller grouped-update atomicity hardening, queue-graph baseline cleanup (wctl check-rq-graph clean with regenerated canonical artifacts), post-enqueue WEPP hint persistence boundary hardening (including lock-contention and non-RuntimeError fault paths with contract-safe behavior), lock/dump-efficiency observability guard coverage, and scoped test maintainability cleanup (shared WEPP payload doubles + less brittle assertions). Required reviewer/qa_reviewer/security_reviewer triad reviews were run after each milestone; all Medium/High findings were remediated before progression. Closure validation passed (228 passed, 0 failed) and enforcement gates were clean (wctl check-rq-graph, check_broad_exceptions --enforce-changed). Post-close addendum documents and remediates a discovered _wepp_bin rollback regression in NoDb Redis cache signature handling (ada260d79).

NoDb Lock/Dump Efficiency Refactor (RQ Engine) (2026-04-25)

Status: ✅ COMPLETE
Link: docs/work-packages/20260425_nodb_lock_dump_efficiency_refactor/
Lifecycle: Backlog -> In Progress -> Done (2026-04-25)
Summary: Closed end-to-end with all scoped rq-engine lock/dump hotspots converted to grouped single-lock mutation flows: wepp_run_payload.py, watershed_routes.py, landuse_routes.py, upload_batch_runner_routes.py, wepp_routes.py, and bootstrap_routes.py, with required NoDb helper additions in Soils, Watershed, Landuse, Disturbed, BatchRunner, and Wepp.persist_job_hint(...). Per-milestone reviewer/qa_reviewer/security_reviewer loops were executed and all Medium findings were remediated before progression; remaining findings are Low-only residual notes. Targeted validation passed across scoped suites (198 passed), and wctl check-rq-graph was executed with documented pre-existing drift while no queue-wiring files were changed in this package. ExecPlan was archived under prompts/completed/.

Landuse/Disturbed MOFE Pipeline Optimization (apprehensive-caw) (2026-04-25)

Status: ✅ COMPLETE
Link: docs/work-packages/20260424_landuse_disturbed_mofe_pipeline_optimization/
Lifecycle: Backlog -> In Progress -> Done (2026-04-25)
Summary: Closed end-to-end with all three prioritized lanes implemented and validated. Lane 1 consolidated duplicate heavy build_managements() work in the DOMLC landuse/disturbed chain via explicit deferred-rebuild contract (Landuse.build() scoped deferral); Lane 2 compacted remap/MOFE hot-loop INFO logging to summary lines while preserving warning/error diagnostics and DEBUG detail; Lane 3 added guarded same-cycle MOFE pair-count reuse with explicit signature checks and invalidation (build-cycle reset + signature drift). Targeted lane regression suites passed (42 passed) across touched landuse/disturbed modules. Required benchmark/parity artifacts were regenerated using isolated temp directories with no source-run mutation under /wc1/runs/ap/apprehensive-caw/; parity status is match for all lanes and benchmark summary includes per-lane baseline/optimized mean/stddev and percent deltas. Code/QA/security review artifacts are closed with no unresolved medium/high findings, and the ExecPlan is archived under prompts/completed/.

Controllers-GL Cache Hardening Rollout (2026-04-24)

Status: ✅ COMPLETE
Link: docs/work-packages/20260424_controllers_gl_cache_hardening/
Lifecycle: Backlog -> In Progress -> Done (2026-04-24)
Summary: Closed end-to-end with include-hygiene hardening for all WEPPcloud templates that load controllers-gl.js. Inventory/remediation covered 19 templates across wepppy/weppcloud/templates/** and wepppy/weppcloud/routes/**; every include now uses static_url('js/controllers-gl.js') and immediately loads controllers_gl_stale_check.js after it while preserving local script order and existing defer/non-defer behavior. Regression coverage was expanded in tests/weppcloud/test_stale_controllers_gl_template_wiring.py to assert inventory-wide invariants. Required validations passed (46 pytest render tests, stale-check Jest suite 9/9, targeted template wiring pytest 26 passed), and both package review gates (artifacts/2026-04-24_code_review.md, artifacts/2026-04-24_qa_review.md) closed with no unresolved medium/high findings.

RQ Worker Startup and NoDb Redis Cache Hardening (Retroactive) (2026-04-24)

Status: ✅ COMPLETE
Link: docs/work-packages/20260424_rq_worker_nodb_cache_hardening/
Lifecycle: Retroactive capture -> Done (2026-04-24)
Summary: Captures completed incident-response hardening for repeated Redis NoDb cache client is unavailable failures impacting landuse mapping jobs. Closed scope includes: safe reconnect semantics for NoDb lock/cache Redis clients, lock-token ownership enforcement before dump() persistence, removal of foreign-lock force-unlock fallback, worker startup gating via docker/rq-worker-startup.sh (readiness probe + configurable delay), compose hardening for prod and worker-only stacks (RQ_REDIS_URL fail-fast contract, REDIS_URL alignment, weppcloudr health dependency), and regression/doc updates. Two review rounds were executed with code/QA/security artifacts; all medium/high findings were dispositioned and targeted validation passed (61 pytest tests, compose config checks, doc-lint clean). Lifecycle-standard alignment is now explicit in package docs (hypotheses/signals, Redis NoDb cache connection strategy, and callus sunset checkpoints), with a post-close observation window through 2026-05-24 and softening review checkpoint on 2026-05-25.

Landuse Legacy Flask State Route Removal (Post Gate 3) (2026-04-24)

Status: ✅ COMPLETE
Link: docs/work-packages/20260424_landuse_legacy_flask_state_route_removal/
Lifecycle: Backlog -> In Progress -> Done (2026-04-24)
Summary: Closed end-to-end with deprecated Flask landuse compatibility machine/state routes removed from wepppy/weppcloud/routes/nodb_api/landuse_bp.py, including set_landuse_mode, set_landuse_db, modify_landuse_coverage, modify_landuse_mapping, user-defined catalog mutators/read route, landuse-map snapshot/save/clear routes, and modify_landuse. WEPPcloud render routes stayed in WEPPcloud (/report/landuse, /landuse-user-defined, /landuse-map) and passed render-route validation. Production in-repo callers no longer target removed Flask endpoints; route/docs/schema ownership now reflects rq-engine-only machine/state APIs. Post-closure remediations include Finder metadata sidecar tolerance for landuse-user-defined/upload, shared page-shell title styling for the landuse catalog/map pages, stale system custom-map reference recovery, a root-cause clean-path fix so build_landuse preserves run-scoped landuse/user-defined/ + landuse/landuse_user_defined_mapping.json, a run_0 render-path recovery boundary so stale system map references do not make projects unloadable (500), a stale-write race closure so stale-system-map cleanup on unlocked reads is in-memory-only with explicit run_0 stale-write retry coverage, custom-map description integrity remediation so changed assignments normalize management labels (for example key 43 surfaces as Moderate Severity Fire) instead of stale base-map descriptions, title-row parity updates so both landuse editor pages expose the run-home runid link left of wc-control__title, and map-description edit/save parity so /landuse-map description edits persist through save/reload/report paths. A final cross-package code/QA/security disposition pass closed additional actionable findings (mapping-selection allowlist hardening, unknown token-class denial, map-path redaction, and landuse-map/save 428/header-body precondition contract parity). Required validation suite passed (22 + 46 + 50 + 54 + 10 pytest results, plus 32 auth-route pytest, 20 + 3 + 2 Jest results, doc-lint clean), and dedicated security review closed with no unresolved medium/high findings.

Landuse Phase 3 Hardening Parity Tests and Migration Gate (2026-04-24)

Status: ✅ COMPLETE
Link: docs/work-packages/20260424_landuse_phase3_hardening_parity_tests/
Lifecycle: Backlog -> In Progress -> Done (2026-04-24)
Summary: Closed end-to-end with Gate 3 pass criteria met. Baseline WEPPcloud hardening behavior is now explicitly frozen for all required matrix rows (path containment, archive/member policy, upload boundaries/conflicts, optimistic concurrency, invalid-row validation, and rollback). Deferred Phase 3 landuse surfaces now have rq-engine parity implementations and tests: user-defined catalog upload/list/delete/update-description, landuse map snapshot/save/clear-override, and modify-landuse. Browser transport for moved surfaces uses session-token bridge bearer flow (requestWithSessionToken) with no cookie-mutation fallback on rq-engine mutators, while render routes remain in WEPPcloud (/report/landuse, /landuse-user-defined, /landuse-map). Required validation suite passed (24 + 39 + 54 + 10 pytest results, 20 + 3 Jest results, doc-lint clean), and dedicated security review closed with no unresolved medium/high findings.

Landuse First-Class Agent Interface Migration (Phased rq-engine Cutover) (2026-04-24)

Status: ✅ COMPLETE
Link: docs/work-packages/20260423_landuse_first_class_agent_interface_migration/
Lifecycle: Backlog -> In Progress -> Done (2026-04-24)
Summary: Closed with phased gate execution and no render-route migration. Gate 0 decisions were finalized before code changes (PUP/active-root strategy, token-class/scope policy, browser transport strategy). Phase 1 shipped first-class rq-engine replacements for set-landuse-mode, set-landuse-db, and modify-landuse-coverage, and WEPPcloud browser callers were cut over to requestWithSessionToken. Phase 2 shipped GET /api/runs/{runid}/{config}/controllers/landuse/state and endpoint-catalog/openapi parity for migrated landuse operations (including existing modify-landuse-mapping discoverability gap closure). Phase 3 map/catalog/file surfaces remained in WEPPcloud by explicit no-go gate policy. Legacy Flask compatibility/deprecation policy is documented with a no-delay removal posture and explicit sunset criteria. Required JS tests passed; required pytest suites passed under .venv/bin/pytest fallback due repeated wctl run-pytest exit 137 in this environment; contract/package docs lint passed.

Landuse User-Defined Management Catalog + Mapping Editor (2026-04-24)

Status: ✅ COMPLETE
Link: docs/work-packages/20260423_landuse_user_defined_management_catalog_map/
Summary: Closed end-to-end with two new PowerUser workflows: Landuse User-Defined catalog and Landuse Map editor. Shipped run-scoped .man/.zip catalog ingest with hardened archive/member/size constraints, metadata CRUD, and all-or-nothing install behavior under landuse/user-defined/. Shipped mapping snapshot/save/clear APIs with optimistic concurrency (X-If-Match-Sha256) and atomic persistence to landuse/landuse_user_defined_mapping.json. Landuse NoDb now prefers configured run-local custom maps via custom_mapping_relpath and raises explicit typed errors for missing/invalid configured custom maps; load_map() now supports explicit JSON mapping paths for run-local resolution. Targeted validations passed (86 passed across touched suites), package doc lint passed (5 files validated, 0 errors, 0 warnings), and dedicated security review closed with no unresolved medium/high findings.

Landuse Batched Mapping Submit UX (Single + Multi-OFE) (2026-04-24)

Status: ✅ COMPLETE
Link: docs/work-packages/20260423_landuse_batched_mapping_submit/
Summary: Closed end-to-end with staged landuse mapping submit UX and batched backend processing. Landuse report mapping selects now stage edits locally with a dedicated secondary apply button and live staged-count messaging; one submit posts canonical mappings[] payload to /rq-engine/api/runs/{runid}/{config}/modify-landuse-mapping. rq-engine now validates + normalizes batch edits (including legacy dom/newdom compatibility), removes mapping depends_on chaining, and enqueues a single mapping job. modify_landuse_mapping_rq now applies normalized batches deterministically under one landuse maintenance-lock window with one completion trigger and pre-mutation validation/rollback safeguards. Post-review disposition closed actionable code/QA findings (lock-gate stale completion short-circuit, null-key validation, readonly/inflight UI guards, and stub parity). Targeted validations passed: microservice route tests (19 passed), RQ mutation-guard tests (23 passed), controller Jest suite (20 passed), and package doc-lint (0 errors).

Landuse Multi-OFE Build Optimization (Lookup/Pass/IO/Logging) (2026-04-23)

Status: ✅ COMPLETE
Link: docs/work-packages/20260423_landuse_multi_ofe_build_optimization/
Summary: Closed end-to-end with multi-OFE landuse optimization in wepppy/nodb/core/landuse.py: SBS burn-remap management-summary reuse, duplicate-pass collapse in Landuse.build() multi-OFE path, explicit first-pass guard behavior (domlc_mofe_d cleared pre-build and MOFE pair-count work skipped when assignment map is missing/empty), and reduced high-volume info logging via compact summaries/debug placement. Added targeted regression coverage for SBS remap parity + lookup reuse, build/event contract preservation, first-pass no-op guard behavior, and logging behavior (14 passed across touched landuse suites). Required benchmark/parity artifacts were regenerated on isolated temp copies for all five required runs; contract parity status is match on all runs.

MOFE .mofe.man Synthesis Process-Pool Migration (2026-04-23)

Status: ✅ COMPLETE
Link: docs/work-packages/20260423_mofe_man_synthesis_process_pool/
Summary: Closed end-to-end with canonical .mofe.man synthesis orchestration in wepppy/nodb/core/landuse.py::_build_multiple_ofe(): spawn-first createProcessPoolExecutor, BrokenProcessPool fork retry, bounded sequential fallback, explicit non-pool exception raising, deterministic hill_<topaz_id>.mofe.man basename validation, and bounded batched worker fan-out (max_workers <= 4). Added targeted landuse tests for success/retry/fallback/non-pool-failure/parity behavior (10 passed) and regenerated required benchmark/parity artifacts on isolated temp copies of the five benchmark runs. Parity matched on all runs (0 mismatches); benchmark deltas on this host remained positive (+143.18%, +443.51%, +34.05%, +63.58%, +286.34%), so the package closes on explicit contract/parity evidence rather than a speedup claim. Review artifacts closed with no unresolved medium/high findings; ExecPlan archived under prompts/completed/.

Multi-OFE Landuse Pair-Count Optimization via wepppyo3 Raster Characteristics (2026-04-23)

Status: ✅ COMPLETE
Link: docs/work-packages/20260423_mofe_landuse_pair_counts_wepppyo3/
Summary: Closed end-to-end with new production API wepppyo3.raster_characteristics.count_intersecting_raster_key_pairs (explicit read/shape failure contracts), canonical release export update (release/linux/py312), and WEPPpy Landuse.build_managements() multi-OFE area path cut over from repeated pairwise np.where scans to Rust one-pass pair counts while preserving area/pct semantics and failure propagation. Validation passed in both repos (cargo test -p raster_characteristics_rust: 2 passed; wepppyo3 raster-characteristics pytest: 5 passed; targeted WEPPpy gates: 9 passed across landuse/daymet/omni suites). Benchmark/parity artifacts were captured on the required five-run matrix in isolated temp dirs: parity match on all runs; timing deltas were -88.67%, +144.42%, -96.34%, -95.41%, -67.56% (the positive delta is the required non-MOFE run objectionable-sublimate, executed with documented synthetic isolated MOFE map + derived single-segment pairing due missing source MOFE inputs).

MOFE Map Migration to wepppyo3 (Topaz Pre-Index + One-Pass Rank Assignment) (2026-04-23)

Status: ✅ COMPLETE
Link: docs/work-packages/20260423_mofe_map_wepppyo3/
Summary: Closed end-to-end with MOFE map assignment production behavior moved from WEPPpy Python loops to wepppyo3.watershed_abstraction.assign_mofe_map (new crate/module), keeping explicit fallback/repair contracts and contiguous-id guarantees. WEPPpy _build_mofe_map now delegates to the Rust path via strict loader wepppy/topo/watershed_abstraction/mofe_map.py; legacy Python assignment remains available only as parity oracle helper (_build_mofe_map_labels_python_legacy) and is not used as silent fallback. Validation artifacts captured on /wc1/runs/po/pointy-toed-fluff subset (200 hillslopes, isolated temp dirs): parity mismatch_count=0; benchmark (6 alternating samples) old mean 65.949408s, new mean 0.282354s, -99.57%. Targeted gates passed in both repos (cargo test -p watershed_abstraction_rust, wepppyo3 pytest, WEPPpy targeted wctl run-pytest). Broad WEPPpy sweep (wctl run-pytest tests --maxfail=1) surfaced one unrelated existing failure in tests/nodb/test_wepp_run_service.py::test_run_watershed_does_not_rewrite_wepp_50k_bin.

Segmented MOFE Migration to wepppyo3 + Process-Pool Refactor (2026-04-23)

Status: ✅ COMPLETE
Link: docs/work-packages/20260422_segmented_multiple_ofe_wepppyo3_pool/
Summary: Closed end-to-end with MOFE segmentation production behavior moved to wepppyo3 (wepp_interchange.segment_single_ofe_slope), WEPPpy SlopeFile.segmented_multiple_ofe hard-switched to Rust path, and legacy Python segmentation retained only behind explicit deprecation (segmented_multiple_ofe_legacy). Refactored WatershedOperationsMixin._build_multiple_ofe to canonical spawn-first process-pool orchestration with fork retry on BrokenProcessPool, bounded sequential fallback, and explicit non-pool failure raising. Validation artifacts captured on /wc1/runs/po/pointy-toed-fluff: parity (3345 files checked, 0 mismatches) and alternating benchmark (old mean 2.148501s, new mean 0.938375s, -56.32%). Targeted test suites passed (wepppyo3: 8 passed; WEPPpy: 17 passed via direct local pytest while wctl execution was blocked by stopped weppcloud compose service). ExecPlan archived under prompts/completed/ with outcome note.

Peridot Side-Hillslope Length Capping + Provenance Mode (2026-04-23)

Status: ✅ COMPLETE
Link: docs/work-packages/20260422_peridot_side_hillslope_length_capping/
Summary: Closed with side-hillslope length capping delivered in both Peridot abstraction paths (L_final = min(L_area, L_edge) for side hillslopes), area-preserving width recomputation, unchanged top/source behavior, and additive hillslope provenance fields (length_estimate_mode, length_area_over_channel, length_edge_median). Peridot tests and schema/manifest checks passed, WEPPpy contract docs were updated, and ExecPlan artifacts were archived under prompts/completed/.

Disturbed MOFE 9002 Soil Support Parity (2026-04-22)

Status: ✅ COMPLETE
Link: docs/work-packages/20260421_disturbed_mofe_9002_soils/
Summary: Completed end-to-end with an explicit MOFE sol_ver=9002 contract locked to single-OFE reference semantics for lookup hits and a documented MOFE-specific lookup-miss deviation required by same-version MOFE stack synthesis. Implemented minimal Disturbed.modify_mofe_soils update so 9002 lookup misses remain class-aware (mukey-texid-disturbed_class) while preserving explicit fallback replacements (luse, stext, ksatfac=0.0, ksatrec=0.0). Added MOFE 9002 regression coverage for lookup-hit, lookup-miss, treatment-suffix normalization, class-keying, and area/pct recomputation. Required gates passed: disturbed single+MOFE tests (17 passed), lookup contract tests (30 passed), soil util tests (49 passed). Config-level check against disturbed9002-10-mofe.cfg confirmed disturbed.sol_ver=9002.0 and wepp.multi_ofe=true flag presence.

Geneva Interactive Summary Report (Retroactive) (2026-04-18)

Status: ✅ COMPLETE
Link: docs/work-packages/20260418_geneva_interactive_summary_report/
Summary: Recorded retroactive closure for the completed Geneva summary report upgrade. Work shipped interactive /query/geneva/summary and /report/geneva/summary contracts (filters/options/chart metadata/selected storm/event table), canonical Pure report rendering, marker-to-table linkage, and review-driven hardening (stale-summary suppression, payload sanitization, no-store response headers). Follow-on runtime fixes ensured _base_report.htm shell dependencies are always provided (ron, current_ron, unitizer_nodb, precisions), and regression coverage was updated accordingly.

Iterative First-Order Link Prune WP-10 WEPPpy E2E Cutover (2026-04-14)

Status: ✅ COMPLETE
Link: docs/work-packages/20260414_ifolp_wp10_wepppy_e2e_cutover/
Summary: Closed WP-10 end-to-end with WEPPpy IFOLP cutover behavior: default stream-pruning method is ifolp, explicit legacy mode remove_short_streams remains selectable, and IFOLP call-site explicitly passes max_junctions=3. Completed watershed/rq-engine/UI method plumbing (state defaults, payload validation, schema-default reporting, controller/template propagation), added method-matrix assertions for both pruning paths, and executed required gates: wctl run-pytest tests/microservices/test_rq_engine_watershed_routes.py (36 passed), wctl run-pytest tests/rq/test_project_rq_mutation_guards.py (11 passed), wctl run-pytest tests/topo/test_terrain_processor_wbt_integration.py (4 passed), wctl run-pytest tests/culverts/test_culvert_batch_rq.py (4 passed), wctl run-npm lint, wctl run-npm test (76 suites, 509 tests). Review disposition closed with no unresolved high/medium findings; ExecPlan archived under prompts/completed/.

Iterative First-Order Link Prune WP-09 Max Junctions Support (2026-04-14)

Status: ✅ COMPLETE
Link: docs/work-packages/20260414_ifolp_wp09_max_junctions_support/
Summary: Closed WP-09 end-to-end with IFOLP --max_junctions support in weppcloud-wbt (Rust tool contract + Phase B behavior + parser/phase tests + both Python wrappers), retained omitted-argument baseline behavior, and deterministic explicit --max_junctions=3 validation. Required gates passed: cargo check -p whitebox_tools; cargo test -p whitebox_tools iterative_first_order_link_prune -- --nocapture (77 passed, 0 failed); python -m py_compile whitebox_tools.py WBT/whitebox_tools.py. Parity/regression artifacts for omitted and max_junctions=3 modes on run1/run2 matched retained canonical hash 920cc1612bd677a1f8dab935a521f6270e226bf961fd5f72ca770b32cd134c83, with no-arg artifacts byte-identical to retained baseline canonicals and max_junctions=3 deterministic across reruns. WEPPpy integration planning now explicitly requires max_junctions=3; review disposition closed with no unresolved high/medium findings; ExecPlan archived under prompts/completed/.

Iterative First-Order Link Prune WP-08 WBT Wrapper Exposure + Release Readiness (2026-04-13)

Status: ✅ COMPLETE
Link: docs/work-packages/20260413_ifolp_wp08_wrapper_release_readiness/
Summary: Closed WP-08 with wrapper-surface release readiness and retained-baseline parity stability. Added iterative_first_order_link_prune to both Python wrapper surfaces (whitebox_tools.py, WBT/whitebox_tools.py), validated CLI/wrapper contract checks (--listtools, --toolhelp=IterativeFirstOrderLinkPrune, missing required args, threshold-pair validation), and passed required gates (cargo check -p whitebox_tools, targeted IFOLP tests 51 passed, 0 failed, wrapper py_compile). Parity spot checks on /tmp/ifolp_wp05_remediate/run1 and run2 produced canonical hash 920cc1612bd677a1f8dab935a521f6270e226bf961fd5f72ca770b32cd134c83, byte-identical to retained parity-report.final_effective.canonical.json artifacts. Review disposition closed with no unresolved high/medium findings.

Iterative First-Order Link Prune WP-07 Optimization Pass (2026-04-13)

Status: ✅ COMPLETE
Link: docs/work-packages/20260413_ifolp_wp07_optimization_pass/
Summary: Closed WP-07 with bounded topology optimization and preserved retained behavior. Performance deltas (5 repeats, run1 fixtures): blackwood_60_5 0.046s -> 0.042s (-8.70%), clueless_aftertaste_anchor_10_100 0.020s -> 0.020s (0.00%), gatecreek_10m_30_2 0.750s -> 0.706s (-5.87%). Parity-regression canonical hash remained 920cc1612bd677a1f8dab935a521f6270e226bf961fd5f72ca770b32cd134c83 on run1/run2 and matched retained baseline artifacts. Required gates passed: cargo check -p whitebox_tools; targeted IFOLP suite passed (51 passed, 0 failed). Review disposition recorded no unresolved high/medium findings.

Iterative First-Order Link Prune WP-06 Error Contract + Robustness Hardening (2026-04-13)

Status: ✅ COMPLETE
Link: docs/work-packages/20260413_ifolp_wp06_error_contract_robustness_hardening/
Summary: Closed WP-06 with bounded hardening only (no pruning-semantic changes). Added explicit non-finite numeric rejection in parser and threshold-table values, duplicate threshold-code detection, and finite/non-negative boundary guards across Phase A/Phase B/topology inputs. Required gates passed: cargo check -p whitebox_tools; cargo test -p whitebox_tools iterative_first_order_link_prune -- --nocapture (50 passed; 0 failed). Parity-regression reruns on /tmp/ifolp_wp05_remediate/run1 and run2 produced stable canonical hash 920cc1612bd677a1f8dab935a521f6270e226bf961fd5f72ca770b32cd134c83 and were byte-identical to retained parity-report.final_effective.canonical.json artifacts (no retained-state drift). No unresolved high/medium findings.

Iterative First-Order Link Prune WP-05 TopAZ Parity Validation (2026-04-14)

Status: ✅ COMPLETE
Link: docs/work-packages/20260413_ifolp_wp05_topaz_parity_validation/
Summary: Closed WP-05 with stakeholder-accepted effective parity equivalence. Retained IFOLP state includes H-002 + H-009 + H-010 + H-011; anchor fixture reached exact parity and non-anchor residuals were accepted low-severity after provenance-aligned probe evidence. Current retained artifact canonical hash is 920cc1612bd677a1f8dab935a521f6270e226bf961fd5f72ca770b32cd134c83 (run1/run2 stable; historical governance token 07e351... retained in older WP-05 records). Final gates passed: cargo check -p whitebox_tools; cargo test -p whitebox_tools iterative_first_order_link_prune -- --nocapture (43 passed; 0 failed). Remaining parity work must treat this retained IFOLP state as the baseline unless a new work package explicitly revises the baseline.

Iterative First-Order Link Prune WP-04 First-Order-Link Pruning (2026-04-13)

Status: ✅ COMPLETE
Link: docs/work-packages/20260413_ifolp_wp04_first_order_link_pruning/
Summary: Completed WP-04 end-to-end in /workdir/weppcloud-wbt with Phase B pruning behavior in companion module iterative_first_order_link_prune_phase_b.rs. Implemented deterministic receiver-group shortest-link selection with strict epsilon improvement, immediate prune mutation (receiver-preserving normal case + self-receiver terminal special case), stale-candidate skip, degeneration-driven repass cadence with deterministic termination, and single-link parity guard behavior. Tool orchestration now executes Phase A -> Phase B and writes final binary output raster + metadata. Required gates passed: cargo check -p whitebox_tools; cargo test -p whitebox_tools iterative_first_order_link_prune -- --nocapture (39 passed; 0 failed). Review findings were dispositioned with no unresolved high/medium issues; ExecPlan archived under prompts/completed/.

Iterative First-Order Link Prune WP-03 Source-Area Qualification (2026-04-13)

Status: ✅ COMPLETE
Link: docs/work-packages/20260413_ifolp_wp03_source_area_qualification/
Summary: Completed WP-03 end-to-end in /workdir/weppcloud-wbt with Phase A source-area qualification behavior in companion module iterative_first_order_link_prune_phase_a.rs. Implemented minimum-CSA provisional masking, row-major inline source-walk mutation, receiver transitions (junction collapse and terminal receiver recheck), and stabilization reclassification, then wired Phase A into tool orchestration while keeping the WP-04 Phase B boundary explicit via unsupported placeholder. Required gates passed: cargo check -p whitebox_tools; cargo test -p whitebox_tools iterative_first_order_link_prune -- --nocapture (33 passed; 0 failed). Review findings were dispositioned with no unresolved high/medium issues; ExecPlan archived under prompts/completed/.

Iterative First-Order Link Prune WP-02 Topology Kernel (2026-04-13)

Status: ✅ COMPLETE
Link: docs/work-packages/20260412_ifolp_wp02_topology_kernel/
Summary: Completed WP-02 end-to-end in /workdir/weppcloud-wbt with deterministic topology-kernel primitives and companion tests. Implemented Whitebox/ESRI pointer decode + neighbor traversal, topology classification + receiver detection, deterministic first-order-link discovery order, and stale-candidate validity checks. Dispositioned review findings (terminal-head half-cell behavior, stream-mask geometry validation for inflow counts, non-negative epsilon clamp + parser rejection of negative epsilon) and expanded synthetic-grid coverage for inflow/state/order/tie/stale behavior. Required gates passed: cargo check -p whitebox_tools; cargo test -p whitebox_tools iterative_first_order_link_prune -- --nocapture (28 passed; 0 failed). WP-02 row in WBT implementation plan marked done with review/test fields complete; ExecPlan archived under prompts/completed/.

Iterative First-Order Link Prune WP-01 Tool Scaffolding (2026-04-13)

Status: ✅ COMPLETE
Link: docs/work-packages/20260412_ifolp_wp01_tool_scaffolding/
Summary: Completed WP-01 tool scaffolding in /workdir/weppcloud-wbt with IFOLP command creation, registry wiring (stream_network_analysis/mod.rs, tools/mod.rs), parser/default/help contract implementation, and targeted parser/registration/placeholder tests (13 passed). Parser tests were split into companion module iterative_first_order_link_prune_parser_tests.rs to keep tool source maintainable and non-monolithic.

Iterative First-Order Link Prune WP-00 Parity Harness (2026-04-13)

Status: ✅ COMPLETE
Link: docs/work-packages/20260412_ifolp_wp00_parity_harness/
Summary: Completed WP-00 end-to-end for IFOLP in /workdir/weppcloud-wbt. Delivered checksum-pinned fixture catalog and oracle manifest, parity metric specification, deterministic rerun report, and reusable harness utilities (ifolp_wp00_prepare_fixtures.py, ifolp_wp00_run_topaz_oracle.sh, ifolp_wp00_compare_outputs.py). Required anchor fixture /wc1/runs/cl/clueless-aftertaste/dem/wbt was included. WP-00 orchestration row in WBT implementation plan is done with review/test/parity gates complete; deterministic canonical hash matched across reruns.

Upload Boundary Helpers Unification (2026-04-12)

Status: ✅ COMPLETE
Link: docs/work-packages/20260412_upload_boundary_helpers_unification/
Summary: Unified non-ZIP upload boundary logic behind canonical helpers in wepppy/microservices/upload_boundary.py and migrated duplicated helper stacks out of wepppy/microservices/rq_engine/ash_routes.py, wepppy/microservices/rq_engine/omni_routes.py, and wepppy/weppcloud/routes/nodb_api/roads_bp.py via upload_helpers.py compatibility paths. Preserved per-endpoint caps/allowlists and status semantics (including explicit 413 oversize mapping), added helper and route parity regressions for extension/size behavior, and documented helper ownership in docs/schemas/upload-endpoint-contract.md. ZIP canonical controls remained anchored to wepppy/microservices/shape_converter/archive_validation.py; culvert semantic validation ownership remained in wepppy/microservices/culvert_payload_validator.py. Validation gates passed (137 targeted tests; full suite 3511 passed, 36 skipped) and the dedicated security artifact closed with no unresolved medium/high findings.

Upload Endpoints Hardening (2026-04-12)

Status: ✅ COMPLETE
Link: docs/work-packages/20260411_upload_endpoints_hardening/
Summary: Completed scoped non-shape_converter upload hardening end-to-end. Culvert ZIP ingestion now reuses validated archive controls from wepppy/microservices/shape_converter/archive_validation.py (read-with-limit, safe member validation, controlled extraction), while keeping culvert semantic payload checks in wepppy/microservices/culvert_payload_validator.py. Added explicit pre-write size/type enforcement for upload_huc_fire, batch-runner geojson+sbs uploads, landuse/treatments user-defined uploads, disturbed SBS uploads, and Roads GeoJSON upload route. Upload-facing error payloads for scoped endpoints no longer leak traceback internals and remain canonical. Regression coverage now includes ZIP abuse fixtures (traversal, encrypted, nested archive, unsupported compression, duplicate paths, quota) and endpoint size/type checks. Validation gates passed (76 targeted tests; full suite 3502 passed, 36 skipped), and the dedicated security review artifact closed with no unresolved medium/high findings.

RQ Operator Experience Hardening (2026-04-11)

Status: ✅ COMPLETE
Link: docs/work-packages/20260411_rq_operator_experience_hardening/
Summary: Completed end-to-end operator hardening across auth bootstrap, revision coherence, freshness semantics, and smoke reliability. Shipped machine-safe bootstrap endpoint POST /weppcloud/api/auth/rq-engine-operator-token (scope-intersection enforcement, pre-revocation throttling, denylist revocation checks with explicit 503+Retry-After outage contract, audit logging, short TTL/no-store response, CSRF exemption for bearer flow), added run_state_domain + phased run_state_vector semantics across run-scoped snapshot reads, enforced explicit freshness semantics (updated_at, data_state, data_updated_at) with revision-coherent/non-future behavior, and updated descriptor/schema contract fields and regression tests. Maintainer preflight gate passed (251 microservice tests + parity guards), API-only operator acceptance evidence was rerun and captured with UTC/redacted logging, and independent reviewer/qa_reviewer/security_reviewer re-reviews closed with no unresolved medium/high findings. Follow-up acceptance remediation on 2026-04-11 closed the build-climate parser blocker (canonical validation_error payloads, no traceback leakage), aligned climate mode schema/defaults (including mode 5/future-year semantics), and added batched run-endpoint discovery via include_operation_docs=true.

RQ Controller State Contract Cutover (2026-04-11)

Status: ✅ COMPLETE
Link: docs/work-packages/20260410_rq_controller_state_contract_cutover/
Summary: Completed row-8 contract freeze/cutover reconciliation across schema/pointer docs, frozen inventory/checklist parity notes, and package lifecycle evidence. Required code gates passed, phased reviewer -> qa_reviewer -> security_reviewer reviews were dispositioned with no unresolved medium/high findings, and one explicit accepted residual/design risk (session-token scope bridge compatibility) was formally documented with owner and follow-up trigger. Post-close readiness hardening added a canonical end-to-end smoke runbook and reran a consolidated 248-test rq-engine smoke baseline plus guard checks.

RQ Controller State Auth and Concurrency (2026-04-10)

Status: ✅ COMPLETE
Link: docs/work-packages/20260410_rq_controller_state_auth_concurrency/
Summary: Hardened controller-state auth/concurrency semantics by shipping rq:read rollout parity, session-token write-precondition enforcement, and idempotency replay/mismatch parity with descriptor/OpenAPI metadata. All required code gates and independent reviewer/qa_reviewer/security_reviewer gates passed; security closeout recorded one accepted residual design risk and no unresolved medium/high defects.

RQ Controller State Errors, Progress, and Outputs (2026-04-10)

Status: ✅ COMPLETE
Link: docs/work-packages/20260410_rq_controller_state_errors_progress_outputs/
Summary: Implemented run-scoped endpoint error catalogs, async progress metadata integration, and GET /api/runs/{runid}/{config}/outputs with artifact trust/provenance metadata and retrieval handles. Required code gates and independent reviewer/qa_reviewer/security_reviewer gates passed with no unresolved medium/high findings.

Lifecycle Corrections (2026-04-10)

Status: ✅ COLUMNS RECONCILED
Description: The following work packages were still listed in Backlog/In Progress but package-level docs already recorded completion. They were moved to Done after subagent review + doc verification.

• docs/work-packages/20260329_features_export_legacy_exports_cutover/ — Closed (2026-03-29)
• docs/work-packages/20260329_features_export_live_run_matrix/ — Completed (Phase 3 closed 2026-04-01)
• docs/work-packages/20260329_features_export_artifact_readme_metadata/ — Closed (2026-03-29)
• docs/work-packages/20260327_roads_point_source_inslope_non_channel/ — Closed - Production Validation Verified (2026-03-28)
• docs/work-packages/20260327_roads_point_source_outslope_rutted/ — Completed (2026-04-07)
• docs/work-packages/20260327_roads_outslope_unrutted_mofe_replacement/ — Complete (2026-04-08)
• docs/work-packages/20260325_rusle_momm2025_r_mode/ — Complete (2026-03-26)
• docs/work-packages/20260403_roads_map_drilldown/ — Complete; ExecPlan archived (implementation completed 2026-04-04, lifecycle verified 2026-04-28)
• docs/work-packages/20260327_roads_peridot_trace_core/ — Complete - Handoff Ready (2026-03-27)
• docs/work-packages/20260323_roads_wepp_reports_regen/ — Completed (Milestones 1-10 closed, 2026-03-24)
• docs/work-packages/20260124_sbs_map_refactor/ — Closed (2026-01-24)
• docs/work-packages/20251028_wojak_lives/ — Closed - Deferred Follow-On (2026-04-10 05:50 UTC)
• docs/work-packages/20260331_wcag21aa_frontend_accessibility/ — Closed (2026-04-10 05:50 UTC)
• docs/work-packages/20260208_rq_engine_agent_usability/ — Closed (2026-04-10 06:08 UTC)

---

RQ Controller State Geospatial and Upload Metadata

Completed: 2026-04-10
Duration: 1 focused session + remediation/re-review loop
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260410_rq_controller_state_geospatial_uploads/
Description: Implemented run geospatial metadata and upload metadata contract hardening so agents can select first-step watershed defaults and validate upload payloads pre-submit.

Outcome:

• Added rq-engine endpoint:
  • GET /api/runs/{runid}/{config}/geospatial-metadata
• Hardened upload descriptor/schema/default metadata for:
  • rq_engine_upload_dem
  • rq_engine_upload_cli
  • rq_engine_upload_sbs
  • rq_engine_upload_cover_transform
• Aligned cross-surface parity for climate/soils mode constraints and watershed defaults.
• Added explicit runtime max_bytes upload enforcement + oversize regression tests for DEM/CLI/SBS/cover-transform handlers.
• Completed reviewer/QA/security re-reviews with no unresolved medium/high findings.
• Closed lifecycle docs, completed security artifact, and archived ExecPlan to prompts/completed/ with outcome note.

Validation Notes:

• wctl run-pytest tests/microservices/test_rq_engine_geospatial_upload_metadata_routes.py --maxfail=1 (21 passed)
• wctl run-pytest tests/microservices/test_rq_engine_openapi_contract.py --maxfail=1 (9 passed)
• python tools/check_endpoint_inventory.py (pass)
• python tools/check_route_contract_checklist.py (pass)
• wctl run-pytest tests/tools/test_endpoint_inventory_guard.py tests/tools/test_route_contract_checklist_guard.py --maxfail=1 (2 passed)

---

RQ Controller State Schema and Defaults

Completed: 2026-04-10
Duration: 1 focused session + review remediation loop
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260410_rq_controller_state_schema_defaults/
Description: Implemented and closed run-scoped controller and endpoint schema/default metadata reads so agents can discover constraints and run-resolved defaults directly from rq-engine.

Outcome:

• Added rq-engine schema/default endpoints:
  • GET /api/runs/{runid}/{config}/controllers
  • GET /api/runs/{runid}/{config}/controllers/{controller}/schema
  • GET /api/runs/{runid}/{config}/controllers/{controller}/hints
  • GET /api/runs/{runid}/{config}/controllers/{controller}/templates
  • GET /api/runs/{runid}/{config}/endpoints
  • GET /api/runs/{runid}/{config}/endpoints/{operation_id}/schema
  • GET /api/runs/{runid}/{config}/endpoints/{operation_id}/defaults
• Added deterministic metadata assembly and route wiring in schema_defaults_routes.py + app registration.
• Resolved reviewer/QA/security findings, including:
  • climate default type parity (climate_mode_code integer defaulting)
  • operation schema/default parity with live handlers
  • disturbed-mod-aware /upload-sbs availability gating
• Package lifecycle closed:
  • tracker/package/security artifact updated
  • ExecPlan archived to prompts/completed/ with outcome note
  • no unresolved medium/high QA or security findings

Validation Notes:

• wctl run-pytest tests/microservices/test_rq_engine_schema_defaults_routes.py --maxfail=1 (43 passed)
• wctl run-pytest tests/microservices/test_rq_engine_openapi_contract.py --maxfail=1 (9 passed)
• python tools/check_endpoint_inventory.py (pass)
• python tools/check_route_contract_checklist.py (pass)
• wctl run-pytest tests/tools/test_endpoint_inventory_guard.py tests/tools/test_route_contract_checklist_guard.py --maxfail=1 (2 passed)

---

RQ Controller State Orchestration Reads

Completed: 2026-04-10
Duration: 1 focused session + review remediation loop
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260410_rq_controller_state_orchestration_reads/
Description: Implemented and closed run-scoped orchestration read APIs so agents can deterministically query pipeline/readiness state and choose next actions without UI heuristics.

Outcome:

• Added rq-engine orchestration endpoints:
  • GET /api/runs/{runid}/{config}/pipeline
  • GET /api/runs/{runid}/{config}/readiness
• Added route/openapi/guard/frozen-artifact parity updates for the two new agent-facing routes.
• Resolved independent reviewer/QA/security findings, including:
  • dedicated RunConfigMismatchError + narrow 404 mapping
  • UTC normalization for naive timestamps
  • deterministic empty-timeline updated_at
  • child-job status/ended-at folding to prevent premature completion in fan-out job trees
• Package lifecycle closed:
  • tracker/package/security artifact updated
  • ExecPlan archived to prompts/completed/ with outcome note
  • no unresolved medium/high QA or security findings

Validation Notes:

• wctl run-pytest tests/microservices/test_rq_engine_orchestration_read_routes.py --maxfail=1 (25 passed)
• wctl run-pytest tests/microservices/test_rq_engine_openapi_contract.py --maxfail=1 (9 passed)
• python tools/check_endpoint_inventory.py (pass)
• python tools/check_route_contract_checklist.py (pass)
• wctl run-pytest tests/tools/test_endpoint_inventory_guard.py tests/tools/test_route_contract_checklist_guard.py --maxfail=1 (2 passed)

---

Roads NoDb Inslope End-to-End Implementation

Completed: 2026-04-10
Duration: Multi-milestone package (implementation + closeout validation)
Status: ✅ CLOSED
Owner: Codex
Link: docs/work-packages/20260323_roads_nodb_inslope_e2e/
Description: Closed phase-1 Roads inslope integration package after explicit rollback validation and package handoff completion.

Outcome:

• Rollback validation captured for mod disable roundtrip (roads.nodb backup/restore hash parity), roads artifact isolation contract, and queue rollback hygiene (no active Roads job, no residual submit/runtime locks).
• Targeted rollback-related tests re-run and passing:
  • tests/weppcloud/routes/test_project_bp.py (set_mod subset)
  • tests/rq/test_roads_rq.py
  • tests/nodb/mods/test_roads_controller.py (roads-scope resource assertion)
• Package docs closed, rollback artifact added, and ExecPlan archived to prompts/completed/.

---

Usersum Header ROLE Filter and Threshold Search Ceiling

Completed: 2026-04-09
Duration: 1 focused session + follow-up fixes
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260408_usersum_role_filter/
Description: Closed usersum role-filter package delivering header ROLE discovery filtering, threshold role-ceiling semantics, nav alignment, spec sync, and source/raw canonical-path security hardening.

Outcome:

• Header ROLE selector shipped with PowerUser/Admin/Root option contracts and selected-role persistence.
• Role filter semantics now use threshold ceilings with explicit unauthorized-ceiling handling (403 API + page error path).
• Discovery/nav now honors selected role ceiling; doc pages self-report min_role under breadcrumbs.
• Security finding SEC-01 closed by canonicalizing /usersum/src and /usersum/raw rel-path handling before manifest visibility checks.
• Package lifecycle closed: package/tracker updated, ExecPlan moved to prompts/completed/ with outcome summary.

Validation Notes:

• wctl run-pytest tests/weppcloud/routes/test_usersum_bp.py tests/weppcloud/test_usersum_template_wiring.py --maxfail=1 (50 passed)
• wctl run-pytest tests/weppcloud/routes/test_usersum_bp.py tests/weppcloud/routes/test_usersum_docs_contracts.py tests/weppcloud/routes/test_usersum_docs_index.py tests/weppcloud/test_usersum_template_wiring.py --maxfail=1 (58 passed)
• wctl doc-lint --path wepppy/weppcloud/routes/usersum/specification.md (1 file validated, 0 errors, 0 warnings)

---

RQ Controller State Contract Foundation

Completed: 2026-04-10
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260410_rq_controller_state_foundation/
Description: Closed the foundation contract package by reconciling identifier model semantics, descriptor invariants, and roadmap dependency clarity against frozen 2026-02-08 route artifacts.

Outcome:

• Updated foundation schema docs:
  • docs/schemas/rq-controller-state-contract.md
  • docs/schemas/rq-engine-agent-api-contract.md
• Dispositioned independent reviewer findings and recorded decisions/progress in:
  • docs/work-packages/20260410_rq_controller_state_foundation/tracker.md
  • docs/work-packages/20260410_rq_controller_state_foundation/prompts/completed/rq_controller_state_foundation_execplan.md
  • docs/work-packages/20260410_rq_controller_state_foundation/prompts/completed/rq_controller_state_foundation_execplan_outcome.md
• Closed package lifecycle documentation and readied direct follow-on packages:
  • 20260410_rq_controller_state_setup_discovery
  • 20260410_rq_controller_state_orchestration_reads
  • 20260410_rq_controller_state_schema_defaults
• Lifecycle recorded: Backlog -> In Progress (2026-04-10 04:08 UTC) -> Done (2026-04-10 04:23 UTC).
• Security review artifact added:
  • docs/work-packages/20260410_rq_controller_state_foundation/artifacts/2026-04-10_security_review.md

Validation Notes:

• wctl doc-lint --path docs/schemas/rq-controller-state-contract.md --path docs/schemas/rq-engine-agent-api-contract.md --path docs/work-packages/20260410_rq_controller_state_foundation/package.md --path docs/work-packages/20260410_rq_controller_state_foundation/tracker.md --path docs/work-packages/20260410_rq_controller_state_foundation/prompts/completed/rq_controller_state_foundation_execplan.md --path docs/work-packages/20260410_rq_controller_state_foundation/prompts/completed/rq_controller_state_foundation_execplan_outcome.md --path docs/work-packages/20260410_rq_controller_state_foundation/artifacts/2026-04-10_security_review.md --path PROJECT_TRACKER.md (pass)

---

RQ Controller State Setup Discovery

Completed: 2026-04-10
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260410_rq_controller_state_setup_discovery/
Description: Implemented non-run-scoped setup-discovery endpoints and contract/test/documentation guardrails so agents can discover valid create configs and setup operation contracts without out-of-band docs.

Outcome:

• Added setup-discovery endpoints in rq-engine:
  • GET /api/configs
  • GET /api/configs/{config}
  • GET /api/endpoints
  • GET /api/endpoints/{operation_id}/schema
  • GET /api/endpoints/{operation_id}/defaults
  • GET /api/endpoints/{operation_id}/errors
• Added route/openapi coverage for auth matrix, strict payload contract checks, not-found taxonomy parity, and canonical handled-500 behavior.
• Updated frozen route artifacts and guard mappings for six new agent-facing setup routes.
• Closed medium/high reviewer + QA + security findings (metadata/runtime parity, error-contract boundaries, auth/test coverage).
• Package lifecycle closed with archived ExecPlan and required security artifact:
  • docs/work-packages/20260410_rq_controller_state_setup_discovery/prompts/completed/rq_controller_state_setup_discovery_execplan.md
  • docs/work-packages/20260410_rq_controller_state_setup_discovery/prompts/completed/rq_controller_state_setup_discovery_execplan_outcome.md
  • docs/work-packages/20260410_rq_controller_state_setup_discovery/artifacts/2026-04-10_security_review.md
• Lifecycle recorded: Backlog -> In Progress (2026-04-10 06:58 UTC) -> Done (2026-04-10 07:29 UTC).

Validation Notes:

• wctl run-pytest tests/microservices/test_rq_engine_setup_discovery_routes.py --maxfail=1 (28 passed)
• wctl run-pytest tests/microservices/test_rq_engine_openapi_contract.py --maxfail=1 (9 passed)
• python tools/check_endpoint_inventory.py (pass)
• python tools/check_route_contract_checklist.py (pass)
• wctl run-pytest tests/tools/test_endpoint_inventory_guard.py tests/tools/test_route_contract_checklist_guard.py --maxfail=1 (2 passed)
• wctl doc-lint --path docs/schemas/rq-controller-state-contract.md --path docs/schemas/rq-engine-agent-api-contract.md --path docs/work-packages/20260208_rq_engine_agent_usability/artifacts/endpoint_inventory_freeze_20260208.md --path docs/work-packages/20260208_rq_engine_agent_usability/artifacts/route_contract_checklist_20260208.md --path docs/work-packages/20260410_rq_controller_state_setup_discovery/package.md --path docs/work-packages/20260410_rq_controller_state_setup_discovery/tracker.md --path docs/work-packages/20260410_rq_controller_state_setup_discovery/prompts/completed/rq_controller_state_setup_discovery_execplan.md --path docs/work-packages/20260410_rq_controller_state_setup_discovery/prompts/completed/rq_controller_state_setup_discovery_execplan_outcome.md --path docs/work-packages/20260410_rq_controller_state_setup_discovery/artifacts/2026-04-10_security_review.md --path PROJECT_TRACKER.md (pass)

---

Run Sync Dashboard Source Token Integration

Completed: 2026-04-01
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260401_run_sync_source_token_integration/
Description: Integrated optional source run token support into Run Sync Dashboard and run-sync backend so private source runs can sync with bearer authentication.

Outcome:

• Added optional source token form field in:
  • wepppy/weppcloud/routes/run_sync_dashboard/templates/rq-run-sync-dashboard.htm
• Added dashboard payload wiring in:
  • wepppy/weppcloud/controllers_js/run_sync_dashboard.js
• Updated rq-engine enqueue payload handling in:
  • wepppy/microservices/rq_engine/run_sync_routes.py
  • optional source_run_token parsing and propagation to run_sync_rq.
• Updated worker auth behavior in:
  • wepppy/rq/run_sync_rq.py
  • wepppy/rq/run_sync_rq.pyi
  • worker now adds Authorization: Bearer <token> headers for aria2c.spec and aria2 requests when token is provided.
• Fixed and tested run-sync status serialization fallback arg indexes for config and source_host.
• Updated docs and queue graph artifacts:
  • docs/run_migration_strategy.md
  • wepppy/rq/job-dependencies-catalog.md
  • wepppy/rq/job-dependency-graph.static.json
  • docs/standards/broad-exception-boundary-allowlist.md
• Completed code/QA review artifacts with no open medium/high findings:
  • docs/work-packages/20260401_run_sync_source_token_integration/artifacts/code_review_findings.md
  • docs/work-packages/20260401_run_sync_source_token_integration/artifacts/qa_review_findings.md

Validation Notes:

• wctl run-pytest tests/microservices/test_rq_engine_run_sync_routes.py tests/rq/test_run_sync_rq.py --maxfail=1 (7 passed)
• wctl run-npm lint (pass)
• wctl check-rq-graph (pass; artifacts refreshed)
• python3 tools/check_broad_exceptions.py --enforce-changed --base-ref origin/master (PASS)
• wctl doc-lint --path docs/run_migration_strategy.md --path wepppy/rq/job-dependencies-catalog.md --path docs/standards/broad-exception-boundary-allowlist.md --path docs/work-packages/20260401_run_sync_source_token_integration --path PROJECT_TRACKER.md (7 files validated, 0 errors, 0 warnings)

---

Admin Run-Scoped Token Minting for Sync and Debug Workflows

Completed: 2026-04-01
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260401_admin_run_token_minting/
Description: Added an admin-only run-scoped token minting workflow (24-hour TTL) in PowerUser Actions for credentialed run sync and debugging.

Outcome:

• Added POST /runs/<runid>/<config>/mint-run-token in wepppy/weppcloud/routes/user.py:
  • requires auth + run authorization + Admin/Root role,
  • issues token_class=service JWT scoped to runs=[runid],
  • fixed TTL 86400 seconds,
  • audiences rq-engine and query-engine,
  • returns canonical payload with Cache-Control: no-store.
• Added admin-only PowerUser "Mint Run Token" card in wepppy/weppcloud/templates/controls/poweruser_panel.htm with mint/copy/status/expiry UX using profile-token styling classes.
• Updated tests:
  • tests/weppcloud/routes/test_user_profile_token.py
  • tests/weppcloud/routes/test_pure_controls_render.py
• Updated docs:
  • docs/dev-notes/auth-token.spec.md
  • wepppy/weppcloud/routes/usersum/weppcloud/getting-started.md
• Completed code/QA review artifacts with medium/high findings resolved:
  • docs/work-packages/20260401_admin_run_token_minting/artifacts/code_review_findings.md
  • docs/work-packages/20260401_admin_run_token_minting/artifacts/qa_review_findings.md

Validation Notes:

• wctl run-pytest tests/weppcloud/routes/test_user_profile_token.py tests/weppcloud/routes/test_pure_controls_render.py --maxfail=1 (49 passed)
• wctl run-pytest tests/weppcloud/routes --maxfail=1 (432 passed)
• wctl doc-lint --path docs/dev-notes/auth-token.spec.md --path wepppy/weppcloud/routes/usersum/weppcloud/getting-started.md --path docs/work-packages/20260401_admin_run_token_minting (5 files validated, 0 errors, 0 warnings)
• python3 tools/check_broad_exceptions.py --enforce-changed --base-ref origin/master (PASS)

---

Usersum Manifest-Driven Docs Engine (GitBook Layout + Vendor + PostgreSQL Search)

Completed: 2026-04-01
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260401_usersum_docs_engine/
Description: Converted usersum into a manifest-driven documentation engine with role-aware visibility, vendor sync support, GitBook-like navigation shell, and PostgreSQL FTS/pg_trgm search while retaining compatibility routes.

Outcome:

• Added machine-readable usersum contracts + generated index pipeline:
  • docs_manifest.yaml, nav_tree.yaml, vendors.yaml
  • generated/docs_index.json
  • tooling: tools/usersum_docs_tool.py (validate, sync-vendors, build-index)
• Added vendor sync for initial weppcloud-wbt scope and committed vendored docs under /usersum/vendor/weppcloud-wbt/....
• Refactored usersum runtime to manifest/index-backed resolution with role-enforced visibility and canonical/doc vendor routes:
  • GET /usersum/doc/<doc_id>
  • GET /usersum/vendor/<vendor_id>/<path:filename>
  • compatibility routes preserved (/usersum/view/*, /usersum/src/*, /usersum/raw/*)
• Implemented GitBook-like shell features:
  • top header search (aligned with theme selector),
  • sticky collapsible nav tree,
  • breadcrumb links,
  • theme-aware sidebar/buttons/scrollbar and full-width usersum shell.
• Implemented PostgreSQL search backend integration (FTS + trigram) with explicit fallback behavior and site-prefix-safe route emission.
• Completed required code and QA review artifacts with all medium/high findings resolved:
  • artifacts/code_review_findings.md
  • artifacts/qa_review_findings.md

Validation Notes:

• PYTHONPATH=/workdir/wepppy python3 tools/usersum_docs_tool.py validate (pass)
• PYTHONPATH=/workdir/wepppy python3 tools/usersum_docs_tool.py validate --require-vendor-files (pass)
• python3 tools/check_broad_exceptions.py --enforce-changed --base-ref origin/master (pass)
• wctl run-pytest tests/weppcloud/routes/test_usersum_bp.py tests/weppcloud/test_usersum_template_wiring.py tests/weppcloud/routes/test_usersum_docs_contracts.py tests/weppcloud/routes/test_usersum_docs_index.py --maxfail=1 (28 passed)
• wctl run-pytest tests/weppcloud/routes/test_usersum_bp.py tests/weppcloud/test_usersum_template_wiring.py --maxfail=1 (23 passed)
• Baseline broad-suite gate during package execution:
  • wctl run-pytest tests --maxfail=1 (2971 passed, 36 skipped)
• wctl run-npm lint (pass)
• wctl run-npm test (pass)

---

Disturbed BD Override + Rosetta WC/FC Recompute

Completed: 2026-04-01
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260401_disturbed_bd_rosetta_wc_fc/
Description: Added disturbed lookup bd override support and an opt-in WEPP advanced option that recomputes top-horizon wp/fc using Rosetta when numeric disturbed bd overrides are present.

Outcome:

• Added canonical disturbed lookup schema change (bd after avke) with blank defaults and additive upgrade coverage.
• Added persisted Soils flag rosetta_wc_fc_from_disturbed_bd_override plus WEPP advanced-options checkbox with exact requested label.
• Wired checkbox serialization/persistence through rq-engine WEPP run/prep routes.
• Implemented strict disturbed bd parsing/validation:
  • empty value = no override,
  • malformed non-numeric text = hard error,
  • numeric bounds = 0.6-2.2 g/cm^3.
• Implemented top-horizon-only bd override + optional top-horizon Rosetta wp/fc recomputation in disturbed soil conversion.
• Completed mandatory reviewer and qa_reviewer passes and resolved all medium/high findings with artifact capture.

Validation Notes:

• wctl run-pytest tests/nodb/mods/disturbed/test_lookup_contract.py tests/nodb/mods/disturbed/test_modify_soils_single_ofe.py tests/nodb/mods/disturbed/test_modify_soils_mofe.py tests/wepp/soils/utils/test_wepp_soil_util.py tests/microservices/test_rq_engine_wepp_routes.py tests/nodb/test_soils_gridded_root_creation.py tests/weppcloud/routes/test_pure_controls_render.py --maxfail=1 (154 passed)
• wctl run-pytest tests/microservices/test_rq_engine_wepp_routes.py tests/microservices/test_rq_engine_soils_routes.py --maxfail=1 (23 passed)
• wctl run-pytest tests --maxfail=1 (2952 passed, 36 skipped)
• wctl run-npm lint (pass)
• wctl run-npm test -- wepp (pass)
• wctl check-test-stubs (pass)
• wctl run-stubtest wepppy.wepp.soils.utils.wepp_soil_util (pass)
• wctl run-stubtest wepppy.nodb.core.soils (pass)

---

Disturbed Panel Modal and Landsoil Lookup UX Contract

Completed: 2026-03-30
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260330_disturbed_panel_modal/
Description: Added a dedicated Disturbed modal in the run-page/report More menu, relocated disturbed lookup actions out of PowerUser, and formalized the base/extended lookup workflow plus docs-link helper contract.

Outcome:

• Added new Disturbed modal template with requested sections:
  • landsoil lifecycle actions (reset base, load extended, delete extended),
  • table-resource selection radios (base/disturbed),
  • explicit modify actions (base, extended, sync base to extended),
  • Help link generated via usersum helper with 📄 affordance.
• Removed disturbed lookup action block and external disturbed-doc link from PowerUser panel.
• Added disturbed task routes:
  • POST .../tasks/delete_extended_land_soil_lookup
  • POST .../tasks/sync_base_to_extended_land_soil_lookup
• Extended disturbed controller wiring for delete/sync actions and lookup-variant UI state refresh.
• Added reusable Jinja helper usersum_doc_link(...) and published canonical developer contract doc at docs/ui-docs/disturbed-panel-ui-contract.md.

Validation Notes:

• python3 wepppy/weppcloud/controllers_js/build_controllers_js.py
• wctl run-npm lint
• wctl run-npm test
• wctl run-pytest tests/weppcloud/routes/test_disturbed_bp.py tests/weppcloud/routes/test_pure_controls_render.py tests/weppcloud/test_jinja_filters.py --maxfail=1 (65 passed)
• wctl run-pytest tests --maxfail=1 (2858 passed, 35 skipped)

---

Features Export Profiles + Provenance Zip Packaging

Completed: 2026-03-28
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260328_features_export_profiles_provenance_zip/
Description: Replaced legacy Features Export defaults with profile-driven UX and standardized all export downloads as zip bundles that include payload outputs plus replay/provenance files.

Outcome:

• Added built-in profiles (post-wepp.yml, prep-details.yml) and run-page profile controls (quick profile buttons + profile-text load).
• Added rq-engine profile resolve endpoint: POST /api/runs/{runid}/{config}/export/features/profile/resolve.
• Refactored service packaging so final artifacts are zip bundles containing payload members, manifest.json, profile.yml, built-in profile files, and README.md.
• Extended manifest payload with profile/provenance relpath fields and bumped features-export cache version marker for packaging contract change.

Validation Notes:

• wctl run-pytest tests/nodb/mods/test_features_export_service.py tests/nodb/mods/test_features_export_manifest.py tests/microservices/test_rq_engine_features_export_routes.py tests/weppcloud/routes/test_pure_controls_render.py tests/weppcloud/routes/test_run_0_openet_admin_gate.py --maxfail=1 (113 passed)
• wctl run-pytest tests/nodb/mods/test_features_export_exporters.py tests/nodb/mods/test_features_export_manifest.py --maxfail=1 (21 passed)
• wctl run-npm test -- features_export (22 passed)

---

Features Export Service Compliance Refactor (4-Phase E2E)

Completed: 2026-03-28
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260328_features_export_service_compliance_refactor/
Description: Closed the QA follow-up four-phase service quality pass for features_export by extracting legacy/carrier collaborators, removing dead wrappers, adding missing strict-required branch tests, and validating end-to-end behavior.

Outcome:

• Added collaborators:
  • wepppy/nodb/mods/features_export/legacy_source_materializer.py
  • wepppy/nodb/mods/features_export/carrier_layer_materializer.py
• Refactored service.py to delegate legacy and carrier source materialization responsibilities to collaborators.
• Added discover_layer_sources(..., skip_vector_relpath=...) support to reuse strict required-source policy for legacy flows without duplicated logic.
• Removed dead wrappers and unused helper code in service (_column_metadata_by_id, _identity_column_token, legacy parquet helpers no longer needed after extraction).
• Added missing strict-required tests (file_missing, unsupported_source_kind, and carrier-path materialization error translation).
• Preserved run-path behavior and counts on baseline smoke run (66 subcatchments, 27 channels).

Validation Notes:

• wctl run-pytest tests/nodb/mods/test_features_export_service.py -k "required_source or discover_layer_sources or materialization_error or ensure_join_key" --maxfail=1 (9 passed)
• wctl run-pytest tests/nodb/mods/test_features_export_planner.py tests/nodb/mods/test_features_export_service.py tests/nodb/mods/test_features_export_exporters.py tests/microservices/test_rq_engine_features_export_routes.py --maxfail=1 (65 passed)
• wctl run-pytest tests/weppcloud/routes/test_pure_controls_render.py -k features_export --maxfail=1 (4 passed)
• wctl run-pytest tests/weppcloud/routes/test_run_0_openet_admin_gate.py -k features_export --maxfail=1 (10 passed)
• wctl run-npm test -- features_export (12 passed)
• python3 tools/check_broad_exceptions.py --enforce-changed --base-ref origin/master (pass; net delta -1)

Features Export Service Quality Refactor (Phased E2E)

Completed: 2026-03-28
Duration: 2 focused sessions
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260327_features_export_service_quality_refactor/
Description: Completed phased quality refactor of wepppy/nodb/mods/features_export/service.py with contract hardening, collaborator extraction, strict required-source enforcement, and full validation/evidence closure.

Outcome:

• Removed hidden identity-key fallback and enforced explicit materialization_error behavior when join-key contracts do not resolve.
• Enforced strict required-source handling on both legacy merge and carrier discovery paths (no warning-only degrade for required missing/unsupported sources).
• Extracted service collaborators:
  • wepppy/nodb/mods/features_export/column_selection.py
  • wepppy/nodb/mods/features_export/cache_rehydration.py
• Expanded service regression tests for required-source failure branches, join-key contracts, and malformed cache-entry fallback behavior.
• Updated wepppy/nodb/mods/features_export/specification.md to lock strict required-source and explicit identity-key contract semantics.
• Completed review artifacts with no unresolved medium/high findings.

Validation Notes:

• wctl run-pytest tests/nodb/mods/test_features_export_planner.py tests/nodb/mods/test_features_export_service.py tests/nodb/mods/test_features_export_exporters.py tests/microservices/test_rq_engine_features_export_routes.py --maxfail=1 (62 passed)
• wctl run-pytest tests/weppcloud/routes/test_pure_controls_render.py -k features_export --maxfail=1 (4 passed)
• wctl run-pytest tests/weppcloud/routes/test_run_0_openet_admin_gate.py -k features_export --maxfail=1 (10 passed)
• wctl run-npm test -- features_export (12 passed)
• python3 tools/check_broad_exceptions.py --enforce-changed --base-ref origin/master (pass; net delta -1)
• wctl doc-lint --path wepppy/nodb/mods/features_export/specification.md (pass)
• wctl doc-lint --path docs/work-packages/20260327_features_export_service_quality_refactor/package.md (pass)
• wctl doc-lint --path docs/work-packages/20260327_features_export_service_quality_refactor/tracker.md (pass)
• wctl doc-lint --path docs/work-packages/20260327_features_export_service_quality_refactor/prompts/completed/features_export_service_quality_refactor_execplan.md (pass)

Roads GeoJSON Attribute Discovery and Mapping UI

Completed: 2026-03-26
Duration: 2 focused sessions
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260326_roads_geojson_attribute_mapping/
Description: Added Roads GeoJSON attribute discovery with explicit mapping controls for design, surface, and traffic, plus user-configurable fallback values (surface_default and traffic_default) and end-to-end validation.

Outcome:

• Upload/config payloads now expose discovered top-level feature-property catalog metadata and persisted mapping state.
• Prepare-stage design eligibility now respects mapping-aware key resolution.
• Run-stage surface/traffic now use mapped-field resolution with explicit fallback values (surface: gravel|paved; traffic: high|low|none).
• Roads UI now supports mapping apply workflow and fallback value selection after upload.
• Regression coverage added/updated across NoDb controller, monotonic segment utility, roads routes, and Roads JS controller.
• Manual run-page E2E confirmed by user: UI mapping flow worked as expected and Roads WEPP run completed successfully.

Validation Notes:

• wctl run-pytest tests/nodb/mods/test_roads_controller.py --maxfail=1
• wctl run-pytest tests/nodb/mods/test_roads_monotonic_segments.py --maxfail=1
• wctl run-pytest tests/weppcloud/routes/test_roads_bp.py --maxfail=1
• wctl run-pytest tests/weppcloud/routes/test_pure_controls_render.py --maxfail=1
• wctl run-npm test -- roads
• wctl run-npm lint
• wctl run-pytest tests --maxfail=1

Disturbed Lookup Hardening and Preservation

Completed: 2026-03-26 (reopen addendum)
Duration: 2 focused sessions
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260325_disturbed_lookup_hardening/
Description: Hardened disturbed lookup CSV persistence against user-edit loss, then reopened to add stale-page lockout/recovery UX and double-submit safeguards while preserving ?pup compatibility.

Outcome:

• Hardened disturbed lookup writes with strict payload validation, duplicate-key rejection, and missing-row guardrails to block partial-table truncation.
• Hardened legacy schema upgrade/read behavior so disturbed_class/texid rows remain readable after upgrade.
• Prevented extended lookup export from clobbering editable lookup CSV (disturbed_land_soil_lookup_extended.csv now separate artifact).
• Updated disturbed CSV editor to dynamic header-driven columns for the full lookup schema.
• Added strict optimistic concurrency (if_match_sha256) for disturbed lookup writes and explicit stale/version-unavailable write-block contracts.
• Added stale-page safeguards in disturbed editor: polling-based stale detection, locked editing on stale state, and explicit Load Current Table / Refresh Page actions.
• Added in-flight save table lock to reduce duplicate-submission/user-confusion paths.
• Added route-side observability events for blocked/committed disturbed lookup writes.
• Completed reviewer + QA subagent passes with artifacts and no unresolved medium/high findings.

Validation Notes:

• wctl run-pytest tests/weppcloud/routes/test_disturbed_bp.py --maxfail=1
• wctl run-pytest tests/nodb/mods/test_disturbed_lookup_persistence.py --maxfail=1
• wctl run-pytest tests/nodb/mods -k disturbed --maxfail=1
• wctl run-pytest tests --maxfail=1
• wctl run-stubtest wepppy.weppcloud.routes.nodb_api.disturbed_bp
• wctl check-test-stubs
• wctl run-npm lint
• wctl run-npm test

Peridot Watershed Parquet + Manifest Integration

Completed: 2026-03-21
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260321_peridot_watershed_parquet_manifest/
Description: Implemented direct Peridot watershed parquet outputs plus flag-aware watershed/README.md manifest generation, then switched WEPPpy to parquet-first ingestion with explicit legacy CSV fallback/migration behavior.

Outcome:

• Peridot now writes watershed/hillslopes.parquet, watershed/channels.parquet, and conditional watershed/flowpaths.parquet for both abstract_watershed and wbt_abstract_watershed.
• Peridot now writes watershed/README.md with execution flags, file manifest, tabular schema summary, and conditional notes.
• WEPPpy now consumes watershed parquet directly for new runs, logs explicit CSV fallback warnings for legacy runs, and keeps migrate_watershed_outputs() functional for old CSV-only projects.
• WEPPpy post-processing now refreshes README manifest/schema sections to reflect final canonical parquet outputs after derived-column normalization.
• Added Rust and pytest coverage for parquet generation, manifest conditionals, direct parquet path, legacy fallback, and migration edge cases.
• Completed real-run verification on /wc1/runs/un/unassailable-sensuousness, including slope sanity check against wepp/runs/p*.slp.

Validation Notes:

• cargo test --test watershed_parquet_manifest -- --nocapture passed (3 passed).
• cargo test --test hillslope_slope_scalar -- --nocapture passed (1 passed).
• wctl run-pytest tests/topo/test_peridot_runner_wait.py passed (11 passed).
• wctl run-pytest tests/tools/test_migrations_parquet_backfill.py -k watershed passed.

RUSLE NoDb + Run-Page UI Integration

Completed: 2026-03-21
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260321_rusle_nodb_ui/
Description: Completed RUSLE Milestones 6-7 with end-to-end NoDb orchestration, async RQ build route, run-header/run-page UI integration, preflight/task wiring, stale invalidation, and full review/QA/validation closeout.

Outcome:

• Added Rusle NoDb facade (wepppy/nodb/mods/rusle/rusle.py) and exports.
• Added async RQ/API flow (build_rusle_rq, POST /api/runs/{runid}/{config}/build-rusle).
• Added disturbed-gated mod toggle + dynamic run-page section rendering with Rusle controls after WEPP.
• Added preflight TaskEnum.build_rusle (🔱) checklist/TOC wiring and staleness invalidation on climate and SBS updates.
• Added focused tests across nodb, rq-engine, WEPPcloud routes/templates/controllers, and preflight checklist logic.
• Synchronized frozen route artifacts/checklists for the new agent-facing endpoint (build-rusle) and updated frozen-route count assertion.

Validation Notes:

• Required gates passed: tests/nodb, tests/weppcloud, npm lint/test, broad-exception enforcement, code-quality observability (observe-only), and full suite (2443 passed, 34 skipped).

RUSLE C Modes Implementation (observed_rap + scenario_sbs)

Completed: 2026-03-21
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260321_rusle_c_modes_implementation/
Description: Completed RUSLE C Milestone 5 by implementing the shared C engine, the locked observed_rap and scenario_sbs modes, auditable run-scoped artifacts, and dedicated review/QA/validation artifacts.

Outcome:

• Added new RUSLE C modules under wepppy/nodb/mods/rusle/: c_formula.py, c_lookup.py, c_manifest.py, c_integration.py.
• Added the runtime lookup substrate: wepppy/nodb/mods/rusle/data/rusle_c_lookup.csv.
• Updated wepppy/nodb/mods/rusle/__init__.py exports with the new C helpers and integration entrypoint.
• Implemented observed_rap with the exact locked contract:
  • fg = clamp(100 - bare_ground_pct, 0, 100)
  • C = exp(-0.04 * fg)
  • neutral canopy/roughness/biomass/consolidation terms
• Implemented scenario_sbs with:
  • DEM-aligned disturbed_class.tif
  • disturbed-family normalization (young forest -> forest)
  • burn-only application for forest, shrub, and tall_grass
  • explicit non-burnable policy enforcement and fail-fast missing-row behavior
• Added targeted tests: test_rusle_c_formula.py, test_rusle_c_lookup.py, test_rusle_c_integration.py.
• Captured package artifacts: artifacts/milestone4_review.md, artifacts/milestone5_qa_review.md, artifacts/final_validation_summary.md.

Validation Notes:

• Passed targeted RUSLE C suite (19 passed).
• Passed broad-exception changed-file enforcement and code-quality observability (observe-only).
• Passed full WEPPpy sanity gate (2429 passed, 34 skipped).

RUSLE POLARIS K Implementation + NRCS Benchmark Harness

Completed: 2026-03-21
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260321_rusle_k_polaris_implementation/
Description: Completed RUSLE K Milestone 4 by implementing polaris_nomograph and polaris_epic, adding deterministic gnatsgo/gssurgo benchmark harness support, and shipping sanity comparison + review/QA artifacts.

Outcome:

• Added K implementation modules under wepppy/nodb/mods/rusle/: k_nomograph.py, k_epic.py, k_reference.py, k_compare.py, k_manifest.py, k_integration.py.
• Updated wepppy/nodb/mods/rusle/__init__.py exports with K integration and comparison entrypoints.
• Added targeted K tests: test_rusle_k_nomograph.py, test_rusle_k_epic.py, test_rusle_k_reference_harness.py, test_rusle_k_compare.py, test_rusle_k_integration.py.
• Implemented benchmark mode precedence contract: gssurgo_kffact -> gnatsgo_kffact -> gssurgo_kwfact -> gnatsgo_kwfact.
• Locked Milestone 0 contracts in manifest/docs: depth support (0_5,5_15 with 5/10 cm weights), EPIC OC conversion (OM/1.724), comparison thresholds defaults, and cfvo deferred scope.
• Captured package artifacts: artifacts/milestone4_review.md, artifacts/milestone5_qa_review.md, artifacts/k_benchmark_comparison_summary.md.

Validation Notes:

• Passed targeted K suite (16 passed).
• Passed broad-exception changed-file enforcement and code-quality observability (observe-only).
• Passed full WEPPpy sanity gate (2410 passed, 34 skipped).

RUSLE Static R + WEPPpyo3 Hyetograph API Migration

Completed: 2026-03-21
Duration: 2 focused sessions
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260320_rusle_r_static_hyetograph_api/
Description: Implemented static R (cligen_static) and shared hyetograph helpers in wepppyo3.climate, migrated WEPPpy climate callsites to canonical outputs, and delivered dedicated review + QA-review artifacts.

Outcome:

• Added new wepppyo3.climate API surface for non-breakpoint/breakpoint hyetograph reconstruction, peak-intensity windows, and static-R from CLI.
• Synced canonical py312 runtime release artifacts under /home/workdir/wepppyo3/release/linux/py312/.
• Migrated in-scope WEPPpy consumers (cligen.py, climate artifact export, interchange fallback, return-period staging) to use canonical peak_intensity_* + duration schema handling.
• Removed breakpoint sentinel intensity behavior and ensured exported artifacts include dur, nullable tp/ip, storm_duration_*, and peak_intensity_10/15/30/60.
• Added regression coverage in tests/climate/test_cligen_peak_intensity_contract.py, tests/nodb/test_climate_artifact_export_service.py, and tests/wepp/interchange/test_utils_phase7.py.
• Added deterministic breakpoint intensity assertions, static-R aggregation invariants, repeated non-breakpoint stability checks, and parquet coalescing-precedence coverage.
• Captured Milestone 4/5 review artifacts and final validation summary under package artifacts/.

Validation Notes:

• Passed: targeted migration tests, Rust tests, broad-exception changed-file enforcement, code-quality observability (observe-only), package/spec/tracker doc lint, and full WEPPpy sanity suite (2392 passed, 34 skipped).

Deliverables:

• ✅ New wepppyo3 hyetograph + static-R API implementation and py312 release sync
• ✅ WEPPpy callsite migration + breakpoint artifact contract upgrades
• ✅ Review/QA/final-validation artifacts and completed ExecPlan

---

RUSLE LS Factor Tooling in weppcloud-wbt

Completed: 2026-03-20
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260320_rusle_ls_factor_wbt/
Description: Implemented a purpose-built RusleLsFactor WhiteboxTools command using locked v1 LS science (Desmet-Govers L, McCool/RUSLE S, DInf default routing), plus end-to-end wepppy integration and manifest provenance.

Outcome:

• Added whitebox-tools-app/src/tools/terrain_analysis/rusle_ls_factor.rs and registered it in terrain-analysis exports + tool manager dispatch.
• Added wrapper methods in both binding files (whitebox_tools.py, WBT/whitebox_tools.py) and verified tool discoverability via --listtools.
• Added WEPPpy LS integration (wepppy/nodb/mods/rusle/ls_integration.py, wepppy/nodb/mods/rusle/__init__.py) and regression tests (tests/nodb/mods/test_rusle_ls_integration.py).
• Finalized LS spec edits for default 304.8 m cap, DEM assumptions, stop-mask routing semantics, and metadata contract.
• Validation passed: WBT build/check/tests + wrapper compile checks, targeted LS integration tests, and full WEPPpy suite (2385 passed, 34 skipped).
• Real-run acceptance on 5 /wc1/runs/* DEMs passed with breached-preprocess workflow, including LS identity (< 2e-5 max absolute error), cap enforcement, and expected fail-fast rejection of unconditioned pit-containing DEMs.

Deliverables:

• ✅ New RusleLsFactor tool + registration + Python bindings
• ✅ WEPPpy LS integration entrypoint and tests
• ✅ Closed package artifacts (package.md, tracker.md, completed ExecPlan)

---

Runtime Path Locks Redis Migration

Completed: 2026-03-17
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260317_runtime_path_redis_locks/
Description: Replaced host-local runtime-path lock files with Redis-backed distributed runtime locks and added command-bar runtime directory lock status/clear operations.

Outcome:

• Migrated runtime lock acquire/release/status/clear behavior to Redis in wepppy/runtime_paths/thaw_freeze.py, including compatibility-safe contention checks and token-safe clear behavior.
• Added command-bar routes and UI commands for runtime directory locks (get directory_locks, clear directory_locks) with canonical 503 error payload handling.
• Updated NODIR_LOCKED guidance to direct operators to :clear directory_locks or wait for TTL expiry.
• Added/updated tests in tests/runtime_paths/test_mutations_thaw_freeze_contract.py and tests/weppcloud/routes/test_command_bar_mcp_token.py.
• Incorporated pre-closure subagent code review + QA review findings before final validation.
• Validation passed: tests/runtime_paths, tests/weppcloud/routes, changed-file broad-exception guard, and full suite (2333 passed, 34 skipped).

Deliverables:

• ✅ Redis-backed runtime lock implementation and helper exports
• ✅ Runtime directory lock command-bar backend + frontend controls
• ✅ Regression coverage for clear/status flows and clear-token safety
• ✅ Closed work-package docs/tracker + completed ExecPlan

---

Omni Contrast Hillslope Re-run Recovery (delete_after_interchange)

Completed: 2026-03-17
Duration: 1 focused session
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260317_omni_contrast_hillslope_rerun/
Description: Added a contrast preflight in run_omni_contrasts_rq that reruns hillslopes (without prep and without interchange) for deduped scenarios referenced by queued contrast runs when delete_after_interchange has removed source hillslope outputs.

Outcome:

• Shipped new preflight helpers in wepppy/rq/omni_rq.py to collect deduped scenario keys, resolve scenario working directories, and rerun Wepp.run_hillslopes() before contrast enqueue fan-out, including scenario cli/slp relpaths back to base runs for existing Omni scenario workspaces.
• Kept existing skip/selection semantics intact by deriving rerun targets from finalized run_ids.
• Added regression coverage in tests/rq/test_omni_rq.py for delete-flag-enabled rerun + dedupe behavior and delete-flag-disabled no-rerun behavior.
• Synced boundary allowlist line anchors for wepppy/rq/omni_rq.py after helper insertion shifted line numbers.
• Validation passed: targeted tests, changed-file broad-exception guard, and full suite (2323 passed, 34 skipped).

Deliverables:

• ✅ run_omni_contrasts_rq rerun preflight for contrast scenarios under delete-after-interchange mode
• ✅ Focused regression tests for rerun gate/dedup/order behavior
• ✅ Package closure artifacts (package.md, tracker.md, completed ExecPlan, project tracker updates)

---

POLARIS NoDb Runs Client for Project-Aligned Raster Layers

Completed: 2026-03-14
Duration: 2 focused sessions
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260313_polaris_nodb_runs_client/
Description: Added a run-scoped, config-driven POLARIS NoDb/mods client that fetches endpoint layers and aligns GeoTIFF outputs to project raster grid contracts.

Outcome:

• Shipped wepppy/nodb/mods/polaris/* with catalog-driven layer selection and default top-horizon sand/clay/bd/om acquisition.
• Added async endpoint/task flow: POST /api/runs/{runid}/{config}/acquire-polaris -> fetch_and_align_polaris_rq.
• Added run-local artifacts under polaris/ (*.tif, manifest.json, generated README.md) with attribution/metadata.
• Added targeted unit + microservice tests including acquire_and_align idempotent skip and force_refresh behavior.
• Verified real-run integration on /wc1/runs/in/insightful-peacock with DEM grid parity checks.
• Full-suite sanity passed: wctl run-pytest tests --maxfail=1 -> 2321 passed, 34 skipped.

Deliverables:

• ✅ NoDb mod + config wiring ([polaris] section in disturbed9002_wbt.cfg)
• ✅ rq-engine route + RQ task + RedisPrep task enum wiring
• ✅ Route-freeze artifact updates (endpoint_inventory_freeze_20260208.md, route_contract_checklist_20260208.md)
• ✅ Work-package closure artifacts and completed ExecPlan

---

Tenerife 2026 Data Ingestion

Completed: 2026-03-12
Duration: 1 focused session after discovery setup
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260312_tenerife_2026_data_ingestion/
Description: Integrated Jonay's 2026 Tenerife refresh by switching Tenerife to a dedicated climate station catalog, validating the new DEM keys, and making the Tenerife soil runtime inventory explicit.

Outcome:

• Added a dedicated Tenerife climate catalog (tenerife_stations.db, tenerife_stations.csv, tenerife_par_files/) plus a repeatable builder script.
• Switched the active Tenerife 25 m and 5 m configs off shared ghcn_stations.db and onto tenerife_stations.db.
• Verified live wmesque2 retrieval for tenerife/136_MDT25_TF and tenerife/MDT05_Tenerife.
• Added Tenerife regression coverage for climate catalog loading, config wiring, and supported soil raster coverage.
• Retired the legacy Tenerife 250 m soil/config branch and template-generation artifacts while keeping tf_soil_10.tif as reference-only inventory.

---

WEPPcloud CSRF Rollout with rq-engine API Compatibility

Completed: 2026-02-24 Duration: 1 day Status: ✅ COMPLETE Owner: Codex Link: docs/work-packages/20260224_weppcloud_csrf_rollout/ Description: Implemented global CSRF protection for WEPPcloud cookie-auth mutation routes while preserving bearer-token compatibility for rq-engine/browse/files third-party and agent clients.

Outcome: Browser mutation routes are CSRF-protected by default with template-driven token propagation, bootstrap forward-auth verify remains explicitly exempt, and rq-engine cookie-path session-token issuance now enforces same-origin while bearer flows remain unchanged.

Deliverables:

• ✅ Artifacts: route classification, exemption register, reviewer findings, code quality review, final validation summary
• ✅ Runtime changes: global CSRFProtect wiring, config toggles, base template CSRF propagation, OAuth disconnect migration, bootstrap exemption wiring
• ✅ Frontend hardening: CSRF bootstrap moved to static/js/csrf_bootstrap.js with dedicated Jest coverage
• ✅ Compatibility hardening: rq-engine session-token same-origin checks for cookie path only
• ✅ Proxy hardening: rq-engine forwarded-origin aliases now require explicit opt-in (RQ_ENGINE_TRUST_FORWARDED_ORIGIN_HEADERS=true)
• ✅ Validation gates executed: required pytest slices, npm http suite, npm csrf_bootstrap suite, code-quality observability, doc-lint
• ✅ check_broad_exceptions --enforce-changed PASS after allowlist line-position synchronization

---

Residual Broad-Exception Closure Finish Line

Completed: 2026-02-24 Duration: 1 day Status: ✅ COMPLETE Owner: Codex Link: docs/work-packages/20260224_residual_broad_exception_finishline/ Description: Closed Debt Project #1 residual broad-exception findings for wepppy/query_engine/app/mcp/router.py and wepppy/weppcloud/app.py with required sub-agent orchestration and validation gates.

Outcome: In-scope unresolved findings reached zero (8 -> 0) while preserving boundary behavior. Non-boundary query-engine parse catches were narrowed, true boundaries were retained and allowlist-synchronized, and required targeted/full pytest gates passed.

Deliverables:

• ✅ Required artifacts: baseline/postfix scanner JSON, baseline inventory, scope resolution matrix, final validation summary
• ✅ Required orchestration: baseline explorer, query-engine worker, weppcloud worker, reviewer, test_guardian
• ✅ Gate results: changed-file broad-exception enforcement PASS (router.py delta -7)
• ✅ Validation: targeted suites PASS (36 passed, 18 passed), full-suite sanity PASS (2107 passed, 29 skipped)

---

Redis Persistence Session Durability and RQ DB9 Deploy Flush

Completed: 2026-02-23 Duration: 1 day Status: ✅ COMPLETE Owner: Codex Link: docs/work-packages/20260224_redis_persistence_session_durability/ Description: Enabled durable Redis defaults in stacks with Redis and introduced explicit deploy-time RQ DB9 flush controls with docs/contract updates for session durability.

Outcome: Redis persistence is now durable by default in dev/prod stacks, RQ job resets are explicit and scoped to DB 9 via deploy controls, and session durability expectations/migration implications are documented.

Deliverables:

• ✅ Runtime durability defaults + env knobs (redis-entrypoint, compose dev/prod wiring)
• ✅ Explicit DB9 flush tooling (scripts/redis_flush_rq_db.sh) + deploy flags (--no-flush-rq-db, --require-rq-redis)
• ✅ Required artifacts: baseline/postfix runtime, deploy flush policy runbook, final validation summary
• ✅ Validation gates: compose renders PASS, targeted pytest PASS, broad-exception enforcement PASS, docs lint PASS
• ✅ Final explorer verification: no remaining high/medium issues

---

Correlation ID Structured Logging End-to-End

Completed: 2026-02-23 Duration: 1 day Status: ✅ COMPLETE Owner: Codex Link: docs/work-packages/20260224_correlation_id_structured_logging/ Description: Implemented canonical correlation_id propagation across weppcloud, rq_engine, query_engine, and rq with X-Correlation-ID ingress/egress behavior, queue metadata continuity, and trace compatibility retention.

Outcome: Correlation ID is generated/accepted at ingress, returned in responses, propagated via enqueue/worker metadata, and mapped into query-engine trace_id without contract breakage. Final explorer review surfaced a Flask direct-enqueue gap that was fixed before closure.

Deliverables:

• ✅ Shared utility module: wepppy/observability/correlation.py
• ✅ Required artifacts: baseline inventory, final flow matrix, validation summary, sample log lines
• ✅ Required orchestration: baseline explorer, workers A-D, final explorer review
• ✅ Gate results: targeted suites PASS, broad-exception changed-file enforcement PASS, code-quality observability PASS
• ✅ Validation: wctl run-pytest tests --maxfail=1 PASS (2086 passed, 29 skipped), wctl check-rq-graph PASS

---

Top Modules Broad-Exception Closure

Completed: 2026-02-23
Duration: 1 day
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260224_top_modules_broad_exception_closure/
Description: Closed broad-exception debt for the top remaining module trees, then completed Milestone 6 residual closure to eliminate the remaining global allowlist-aware findings.

Outcome: Initial package scope reached zero unresolved (354 -> 0), and Milestone 6 closed residual global unresolved findings to zero (51 -> 0). Global bare-exception remained zero and full-suite sanity passed after the Milestone 6 refactor pass.

Deliverables:

• ✅ Required artifacts: baseline/post scanner JSON, full module resolution matrix, final validation summary
• ✅ Required orchestration: baseline explorer, Workers A-E with ownership split, final explorer regression review
• ✅ Milestone 6 artifacts: milestone_6_residual_baseline.json, milestone_6_resolution_matrix.md, milestone_6_postfix.json, milestone_6_final_validation_summary.md
• ✅ Gate results: hard bare gate PASS, target/global unresolved gates PASS, changed-file enforcement PASS
• ✅ Validation: wctl run-pytest tests --maxfail=1 PASS (2066 passed, 29 skipped)

---

NoDb Broad-Exception Boundary Closure

Completed: 2026-02-23
Duration: 1 day
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260223_nodb_broad_exception_boundary_closure/
Description: Comprehensive broad-exception closure for wepppy/nodb/** with required sub-agent orchestration, tests-first characterization, narrowing/removal of non-boundary broad catches, and residual boundary allowlist synchronization.

Outcome: NoDb unresolved broad findings in allowlist-aware mode reached zero, NoDb bare except stayed zero, required NoDb/full-suite gates passed, and closure artifacts were published.

Deliverables:

• ✅ Required artifacts: baseline/final scanner JSON, full resolution matrix, final validation summary
• ✅ Required sub-agent orchestration: baseline explorer, workers A/B/C, final explorer review
• ✅ Gate results: hard bare gate PASS, allowlist-aware unresolved gate PASS, changed-file enforcement PASS
• ✅ Validation: wctl run-pytest tests/nodb PASS (501 passed, 3 skipped), wctl run-pytest tests/nodir PASS (135 passed), wctl run-pytest tests --maxfail=1 PASS (2066 passed, 29 skipped)

---

Bare Exception Zero Closure and Boundary Safety (Phase 2)

Completed: 2026-02-23
Duration: 1 day (Phase 2 closure window)
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260223_bare_exception_zero/
Description: Reopened and completed broad-exception boundary closure for weppcloud/routes, rq_engine, and rq after the original bare-exception closure.

Outcome: Target-module unresolved broad findings in allowlist-aware mode reached zero, global bare-exception count remained zero, and final full-suite validation passed on post-fix state.

Deliverables:

• ✅ Required Phase 2 artifacts: baseline/postfix scanner JSON, classification report, final validation summary
• ✅ Required sub-agent orchestration: baseline explorer, 3 subsystem workers, tests/contracts worker, final explorer review
• ✅ Gate results: hard bare gate PASS, target unresolved gate PASS, changed-file enforcement PASS
• ✅ Validation: wctl run-pytest tests --maxfail=1 PASS (2060 passed, 29 skipped)

---

Broad Exception Elimination and Boundary Contract Hardening

Completed: 2026-02-23
Duration: 2 days
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260222_broad_exception_elimination/
Description: Phased elimination of broad exception handlers in runtime-critical production paths with subagent-driven refactoring/review/testing and contract-safe regression controls.

Outcome: Package completed through Milestone 7 with changed-file broad-catch guard activation, documented approved boundaries, and full-suite validation.

Deliverables:

• ✅ Broad-catch reduction from 1120 to 1103 (bare-except: 96 to 82)
• ✅ Milestone artifacts for phases 0-6 plus final closeout summary
• ✅ Changed-file enforcement mode in tools/check_broad_exceptions.py (--enforce-changed, --base-ref)
• ✅ Checker regression coverage including except*/TryStar handling
• ✅ Required full gate: wctl run-pytest tests --maxfail=1 (2048 passed, 29 skipped)
• ✅ Boundary allowlist with owner/rationale/expiry in package artifacts

---

Cross-Service Auth Token Integration Hardening

Completed: 2026-02-19
Duration: 1 day
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260219_cross_service_auth_tokens/
Description: Established one executable cross-service token contract across WEPPcloud, rq-engine, browse, and query-engine MCP with matrix-driven integration and lifecycle validation.

Outcome: Portability, renewal fallback, revocation propagation, rotation overlap/retirement, and grouped/composite runid cookie round-trip are now explicitly tested and mapped to compatibility matrix rows.

Deliverables:

• ✅ Integration harness in tests/integration/ with shared Redis/JWT fixtures
• ✅ Matrix-driven portability tests (MX-A*)
• ✅ Lifecycle integration tests (MX-L*) including grouped cookie round-trip (MX-L4)
• ✅ Auth primitive unit-gap coverage (exp/nbf/iat/leeway, roles/run auth actor helpers)
• ✅ Synced package docs/artifacts (tracker.md, matrix, lifecycle results, ExecPlan closeout notes)

---

Error Schema Standardization (RQ API Migration)

Completed: 2026-01-12
Duration: 2 days
Status: ✅ COMPLETE
Owner: Codex
Link: docs/work-packages/20260111_error_schema_standardization/
Description: Standardized rq-engine and rq/api responses with canonical keys and status-code-first errors, removing legacy aliases.

Outcome: Contract published, legacy keys removed, job polling updated for 404 not_found, tests/docs aligned.

Deliverables:

• ✅ observed-error-schema-usages report
• ✅ rq-response contract documentation
• ✅ canonical error payloads with 4xx/5xx semantics
• ✅ jobstatus/jobinfo 404 polling updates
• ✅ updated tests and documentation

---

VS Code Theme Integration

Completed: 2025-10-29
Duration: 2 days
Status: ✅ COMPLETE
Owner: GitHub Copilot + gpt-5-codex (Codex)
Link: docs/work-packages/20251027_vscode_theme_integration/
Description: Implemented configurable VS Code theme integration to satisfy stakeholder demands for "more style" while preserving zero-aesthetic development philosophy.

Outcome: Configurable theme mapping system delivered with 11 production themes, WCAG AA compliance validation, and stakeholder self-service color editing capability. System unblocks frontend modernization by addressing visual customization concerns without developer burden.

Deliverables:

• ✅ Configurable theme-mapping.json with semantic variable mappings
• ✅ Dynamic converter script with validation and reset capabilities
• ✅ 11 production themes (Light/Dark defaults + 9 VS Code themes)
• ✅ WCAG AA compliance validation for all shipped themes
• ✅ User persistence (localStorage + cookie fallback)
• ✅ Theme switcher UI integrated into settings panel
• ✅ Documentation: theme system guide, stakeholder editing guide, troubleshooting
• ✅ Build pipeline integration with automatic theme generation

---

UI Style Guide Refresh

Completed: 2025-10-27
Duration: 2 days
Status: ✅ COMPLETE
Link: docs/work-packages/20251027_ui_style_guide_refresh/
Description: Merged UI documentation into single agent-training guide with pattern catalog for rapid control construction.

Outcome: Comprehensive pattern catalog delivered with 8 copy-paste templates enabling <5 minute control creation. GPT-5-Codex completed review and validated technical accuracy. Work package handoff completed successfully.

Deliverables:

• ✅ Merged ui-style-guide.md (1151 lines)
• ✅ Pattern Catalog (8 templates)
• ✅ Quick Reference Tables, Troubleshooting, Testing Checklist
• ✅ "Zero-Aesthetic" design philosophy integration
• ✅ GPT-5-Codex technical validation complete

---

Smoke Tests & Profile Harness

Completed: 2025-10-27
Duration: Initial implementation phase complete
Status: ✅ SCOPE COMPLETE
Link: docs/work-packages/20251023_smoke_tests/
Description: Established Playwright-based smoke harness with YAML profile support for health snapshots.

Outcome: Core infrastructure complete and functional. Test-support blueprint operational, smoke harness spec documented, initial profile authored. Scope achieved for immediate needs.

Deliverables:

• ✅ Playwright smoke harness setup
• ✅ YAML profile structure defined
• ✅ Test-support blueprint with SMOKE_RUN_ROOT support
• ✅ Initial quick profile drafted

Note: Future expansion (additional profiles, wctl run-smoke loader) can be addressed in separate work packages as needed.

---

Frontend Integration & Smoke Automation

Completed: 2025-10-27
Duration: ~4 weeks
Status: ✅ SCOPE COMPLETE
Link: docs/work-packages/20251023_frontend_integration/
Description: Completed Pure template migrations, refactored bootstrap initialization, established repeatable smoke validation flow.

Outcome: All primary objectives achieved. Controllers migrated to Pure templates with StatusStream, bootstrap refactored to helper-driven patterns, URL construction standardized, 7 remaining polish issues fully documented for future refinement.

Deliverables:

• ✅ All controllers migrated to Pure templates with StatusStream
• ✅ Bootstrap refactored to helper-driven initialization
• ✅ Map race condition and preflight script issues resolved
• ✅ URL construction pattern (url_for_run()) fixed and documented
• ✅ Seven outstanding polish issues analyzed with implementation specs

Note: Seven polish items (legend styling, table standardization, TOC indicators, map layer wiring, inline help icons, hint deduplication) documented as future enhancements but not blocking production use.

---

NoDb ACID Transaction Update

Completed: 2025-10-25
Duration: 1 day (planning only)
Status: ❌ CANCELED - Unviable
Link: docs/work-packages/20251024_nodb_acid_update/
Description: Proposed Redis-backed ACID transactions for NoDb controllers with intelligent event-driven cache invalidation.

Outcome: After architectural review, the proposed specification was deemed unviable. The approach introduced excessive complexity without sufficient benefit. NoDb's existing file-based state management with Redis locking remains the architectural pattern.

Lessons Learned:

• Redis transactions don't provide the durability guarantees needed for NoDb's file-first architecture
• Cache invalidation rules added complexity without addressing core concurrency patterns
• File-based state with explicit locking is simpler and more maintainable
• Future caching improvements should focus on read-through patterns, not transaction wrappers

---

StatusStream Telemetry Cleanup

Completed: 2025-10-23
Duration: 1 day
Link: docs/work-packages/20251023_statusstream_cleanup/
Description: Replaced legacy WSClient shim with unified controlBase.attach_status_stream helper.

Outcome: Unified telemetry pipeline with no WSClient references; all controllers use StatusStream.

---

Controller Modernization Documentation Backlog

Completed: 2025-10-23 Duration: 1 week
Link: docs/work-packages/20251023_controller_modernization/
Description: Consolidated controller modernization documentation after WSClient removal and helper-first migration.

Outcome: Authoritative helper-first documentation established; archived per-controller plans grouped within work package.

---

🔮 Vision / Long-Term Initiatives

High-level initiatives that haven't been broken down into concrete work packages yet. These represent strategic directions or large efforts requiring planning.

[Vision Item Template]

Proposed: YYYY-MM-DD
Sponsor: [Team or person championing this]
Strategic Value: [Why this matters long-term]
Dependencies: [What needs to happen first]
Next Steps: [What scoping work is needed before creating packages]

Kubernetes Migration

Proposed: 2024-Q3
Sponsor: DevOps
Strategic Value: Enable horizontal scaling, improve deployment automation, support multi-tenant scenarios
Dependencies: Static build process finalization, Redis keyspace configuration
Next Steps: Create scoping package to enumerate migration steps, identify risks, and break into implementable chunks

---

WEPP Model Validation Framework

Proposed: 2025-Q2
Sponsor: Research team
Strategic Value: Systematic validation against field data, improved model confidence, publication-ready metrics
Dependencies: Standardized output formats, validation dataset curation
Next Steps: Gather requirements from hydrologists, survey existing validation approaches, prototype validation metrics

---

📊 Metrics

Cycle Time

Track how long packages take from start to completion:

• Target: 2-4 weeks for most packages
• Current average: [Calculate from recent completions]

Work in Progress

• Current: 3 packages
• Target: 2-4 packages
• Status: ✅ Within target

Throughput

Packages completed per month:

• October 2025: 7 packages completed/closed (6 completed successfully, 1 canceled as unviable); 1 package advanced to Phase 3 complete (markdown-doc toolkit); 3 packages started (UI Style Guide Refresh, VS Code Theme Integration, markdown-doc toolkit)

Lead Time

Time from package creation to completion:

• Track in package tracker.md timeline sections

---

📝 Notes

When to Update This Tracker

Agents should update this tracker when:

• Starting a new work package (add to Backlog or In Progress)
• Moving a package between columns (Backlog → In Progress → Done)
• Significant progress on an active package (update "Recent Progress")
• Blocking issues that affect package status
• Package completion (move to Done, add outcome summary)

Frequency: Check and update during each work session that touches work packages.

Archive Policy

Completed packages are moved from "Done" to "History" section immediately upon completion. This keeps the tracker focused on active work while preserving completion history for reference.

Questions or Issues

If this tracker format isn't working or you have suggestions:

1. Create a work package for "PROJECT_TRACKER improvements"
2. Document specific pain points and proposed solutions
3. Experiment with changes and gather feedback

---

📚 History

October 2025

• ✅ VS Code Theme Integration (completed 2025-10-29) - Configurable mapping system with 11 production themes, WCAG AA compliance
• ✅ UI Style Guide Refresh (completed 2025-10-27) - Pattern catalog with 8 templates enabling <5min control creation
• ✅ Smoke Tests & Profile Harness (completed 2025-10-27) - Playwright harness setup with YAML profile support
• ✅ Frontend Integration & Smoke Automation (completed 2025-10-27) - Pure migration complete, 7 polish items documented
• ✅ StatusStream Telemetry Cleanup (completed 2025-10-23) - Unified telemetry pipeline, WSClient removed
• ❌ NoDb ACID Transaction Update (canceled 2025-10-25) - Unviable architecture, file-first approach retained
• ✅ Controller Modernization Documentation Backlog (completed 2025-10-23) - Helper-first docs established

February 2026

• ✅ NED1 VRT Alignment Audit + Correction (completed 2026-02-05) - Audit script + corrected VRT + USGS report delivered

[Month YYYY]

• [Package name] (completed YYYY-MM-DD) - [One line outcome]

---

🔧 Tracker Maintenance

Last reviewed: 2026-04-10
Next review: 2026-05-10 (monthly)

Review checklist:

• Move stale Done items to History
• Update WIP count and check against limits
• Review In Progress packages for stalls (>6 weeks)
• Verify Backlog priorities still align with current goals
• Reconcile stale lifecycle state transitions in tracker columns
• Update metrics section with recent data