Releases: littlebearapps/pitchdocs
Releases · littlebearapps/pitchdocs
v2.2.0
2.2.0 (2026-03-15)
Added
- full plugin review fixes — version sync, hook exit codes, stale references, eval coverage (48dc592)
- remove out-of-scope skill-authoring skill (44ab591), closes #40
Fixed
- add trigger phrases to feature-benefits skill for better activation (12c4db2)
- correct Outlook Assistant logo URL extension (.svg → .png) (36e208f)
- resolve CI failures — spelling, broken links, and llms.txt refs (e787731)
Documentation
- refresh launch artifacts and bump guide versions for v2.1.0 (#42)
- sync AI context files with context-updater agent and context-quality rule
- sync bridge files with corrected agent and rule counts
v2.1.0
2.1.0 (2026-03-12)
Added
- add per-project activation for advisory features (fc57dd2)
Fixed
- add --plugin-dir to activation eval script for CI compatibility (45866e7)
- correct last 2 activation eval test expectations (cb3af25)
- correct test expectations for activation evals (8d3c20a)
- detect API key auth failures in activation eval pre-flight (1ddc554)
- detect billing and rate limit errors in activation eval pre-flight (ed6e8cc)
- install plugin via CLI marketplace for CI activation evals (76156bd)
- reclassify 2 activation eval tests after failure investigation (c87603d)
- register plugin in installed_plugins.json for CI activation evals (c371239)
- use parent directory for --plugin-dir in activation evals (9fe0ded)
Documentation
v2.0.0
2.0.0 (2026-03-11)
⚠ BREAKING CHANGES
- /pitchdocs:ai-context and /pitchdocs:context-guard moved to ContextDocs (https://github.com/littlebearapps/contextdocs). Both remain as stub redirects. 39 files changed, 16 skills, 3 rules, 1 hook.
Added
- add 3 test artifacts — hook unit tests, banned phrase checker, results doc (dc54272)
- add 6 CI checks — spell check, actionlint, frontmatter validation, llms.txt consistency, orphan detection, token budgets (e2eb70b)
- add CI integration — hook tests, banned phrases, activation eval workflow (45372e4)
- add skill activation eval runner for claude -p testing (2be7a4d)
- split AI context management into ContextDocs plugin (6320b98)
Fixed
- AU English spelling, typos config for brands/HTML, lychee exclude canva.com (8871d20)
- correct hook file path in llms.txt from gitignored .claude/hooks/ to tracked hooks/ (dd6bce0)
- fix activation eval runner — stream-json format, remove tool restriction (7c35f1c)
- heading level increment and dead discussions link (55fcb22)
- resolve markdown lint errors in RESULTS.md and tutorial (5147950)
- resolve shellcheck SC2012 and SC2129 in CI workflows (81f6411)
- revert premature version bump so release-please proposes v2.0.0 (47ad6b5)
Performance
- slim doc-standards.md auto-loaded rule (~200 token reduction) (6f7f9d2), closes #36
- split 4 over-budget skills into companion reference files (ed0cc84)
Documentation
- add activation eval reminder to release workflow (c5cf5dd)
- add Testing & Validation section to CLAUDE.md (9734907)
- add tutorial, launch artifacts, ROADMAP, and rewrite docs hub (a222d6c)
- add Untether integration guide and consolidate cross-references (c1a4d89)
- add V5 activation eval results (80%) and fix V6 name mismatch (61d75d9)
- regenerate llms-full.txt for v2.0.0 (c63a87f)
- update CHANGELOG and README for v1.20.0, add V6 eval results (55e9cce)
- update CHANGELOG.md with skill activation eval and CI check additions (9ae707b)
- update guides for v2.0.0 — remove stale context references (7c54cab)
- update RESULTS.md with Phase 3.5 token budget fix outcomes (8d48cab)