@@ -47,6 +47,11 @@ It is built for researchers and engineers who want to compare LLM-backed, symbol
 
 ## News
 
+- **2026-08**: 🚀 **NetOpsBench v0.2.0** - Large-topology benchmark release.
+  - Add Xlarge CLOS and Fat-tree K=8/K=12 profiles, with 70 generated cases per large topology.
+  - Replace per-client Python Pingmesh and iperf processes with the native Rust client agent for Pingmesh and background traffic.
+  - Harden large-topology fault injection, recovery, observability, and exact runtime teardown.
+  - Publish a versioned DeepSeek validation snapshot across all seven supported scales. See the [v0.2.0 release notes](docs/content/docs/releases/v0.2.0.mdx).
 - **2026-05**: 🎉 **Initial Release** - NetOpsBench is now available as an open arena for agentic network troubleshooting.
   - Provide public SDK with `run_scenario()` and `run_suite()` APIs to launch live network environments from Python.
   - Equip native MCP tools of complete observability utilities and pre-configured SONiC-VS network covering XS, Small, Medium and Large scales.
@@ -96,11 +101,21 @@ Scenario YAML files define the benchmark case: topology scale, traffic profile,
 
 NetOpsBench reports detection, fault type, device/interface localization, runtime, tool calls, and token usage so troubleshooting quality and operational cost can be compared together.
 
-![Composite benchmark score](docs/public/assets/benchmark/fig_avg_score.png)
+| Scale | Topology | Switches | Clients | Cases |
+|---|---|---:|---:|---:|
+| XS | CLOS | 4 | 2 | 14 |
+| Small | CLOS | 6 | 8 | 15 |
+| Medium | CLOS | 12 | 16 | 28 |
+| Large | CLOS | 20 | 64 | 52 |
+| Xlarge | CLOS | 144 | 128 | 70 |
+| Fat-tree K=8 | Fat-tree | 80 | 128 | 70 |
+| Fat-tree K=12 | Fat-tree | 180 | 144 | 70 |
 
-Read [Benchmark Methodology](docs/content/docs/run-benchmarks/methodology.mdx) for scoring definitions and [Benchmark Results](docs/content/docs/run-benchmarks/results.mdx) for an example completed suite.
+![NetOpsBench v0.2.0 DeepSeek validation](docs/public/assets/benchmark/fig_deepseek_v02_quality.png)
 
-Public agent trajectory artifacts are available in the [NetOpsBench Trace Dataset](https://huggingface.co/datasets/yyyyyt/netopsbench-trace), including Harbor/ATIF traces, run reports, and summary CSVs for reproducible analysis.
+Read the [v0.2.0 release notes](docs/content/docs/releases/v0.2.0.mdx), [Benchmark Methodology](docs/content/docs/run-benchmarks/methodology.mdx), and [Benchmark Results](docs/content/docs/run-benchmarks/results.mdx) for the full validation snapshot and scoring definitions.
+
+The public [NetOpsBench Trace Dataset](https://huggingface.co/datasets/yyyyyt/netopsbench-trace) contains both the earlier cross-model snapshot and the [v0.2 seven-scale release](https://huggingface.co/datasets/yyyyyt/netopsbench-trace/tree/main/releases/netopsbench-0.2): 319 validated DeepSeek Harbor/ATIF trajectories across XS through Fat-tree K=12. Aggregate metrics and immutable publication provenance are recorded in the [v0.2 result snapshot](docs/public/assets/benchmark/deepseek_v02_release.json).
 
 ## Learn More
 
 
@@ -120,6 +120,8 @@ or deleted.
 
 ## Cleanup
 
+Xlarge and Fat-tree labs can consume substantial CPU, memory, container, and telemetry capacity. Start with one worker, inspect `docker stats`, host memory and pressure, and confirm one complete baseline before increasing concurrency. Keeping a topology warm is useful for repeated experiments, but do not leave unused large workers running alongside a new campaign.
+
 SDK-visible runtimes:
 
 ```bash
@@ -129,6 +131,8 @@ netopsbench runtime teardown <runtime-name>
 netopsbench runtime teardown --all
 ```
 
+Prefer the exact runtime name. `teardown --all` is intended only when every runtime registered in the current workspace is owned by the caller; it is not a substitute for host-wide Docker cleanup. Do not use global `docker system prune`, broad container-name filters, or shared-network deletion on multi-user hosts.
+
 Manual deployment teardown:
 
 ```bash
 
@@ -5,6 +5,10 @@ description: Documentation map for implementing, running, and evaluating NetOpsB
 
 NetOpsBench evaluates troubleshooting agents on generated data-center network fault scenarios. A run provisions a SONiC-VS / Containerlab topology, injects a controlled fault or healthy episode, exposes runtime evidence, calls an agent, and scores the returned `DiagnosisResult` against scenario ground truth.
 
+<Callout title="NetOpsBench v0.2.0 · 2026-08" type="info">
+The current release adds Xlarge CLOS, Fat-tree K=8/K=12, and the native Rust Pingmesh/traffic client agent. Read the [v0.2.0 release notes](/docs/releases/v0.2.0).
+</Callout>
+
 These docs are organized around the agent-development workflow.
 
 ## Main path
@@ -72,3 +76,4 @@ These docs are organized around the agent-development workflow.
 - [System Overview](/docs/architecture/system-overview) explains the runtime loop, evidence path, worker isolation, and report aggregation.
 - [Benchmark Methodology](/docs/run-benchmarks/methodology) defines scenario coverage, scoring, negative samples, and optional semantic fault-type matching.
 - [Benchmark Results](/docs/run-benchmarks/results) records one completed cross-model run for comparison context.
+- [NetOpsBench v0.2.0](/docs/releases/v0.2.0) summarizes the Xlarge, Fat-tree, native client-agent, and lifecycle release.
@@ -5,6 +5,7 @@
     "quickstart",
     "build-your-agent",
     "run-benchmarks",
+    "releases",
     "debug-operate",
     "extend-netopsbench",
     "architecture"
 
@@ -5,6 +5,8 @@ description: The shortest path from a fresh clone to one completed NetOpsBench r
 
 This page verifies that the host can run one benchmark case. It stops after the first `BenchmarkReport`; agent implementation and larger benchmark runs are covered separately.
 
+Use XS for installation verification even when the eventual target is Xlarge or Fat-tree. It exercises the same SDK, native client-agent, observability, fault, scoring, and cleanup path with a much smaller resource footprint.
+
 ## Requirements
 
 <Callout title="Linux host required" type="warn">
@@ -83,6 +85,8 @@ PYTHONPATH=. python examples/01_run_scenario.py --vendor openai
 
 The run provisions an XS topology, starts observability, injects one generated fault, calls the reference agent, scores the returned diagnosis, and writes a `BenchmarkReport`.
 
+`netopsbench benchmark prepare` only generates topology and scenario assets. It does not deploy a lab, start telemetry, or reserve resources; deployment begins when a session or runtime command is executed.
+
 ## After success
 
 - Open saved agent trajectories with `netopsbench trace view`; it syncs trace-enabled runs into the local Harbor viewer cache automatically.
 
@@ -0,0 +1,4 @@
+{
+  "title": "Releases",
+  "pages": ["v0.2.0"]
+}
@@ -0,0 +1,32 @@
+---
+title: NetOpsBench v0.2.0
+description: Xlarge CLOS, Fat-tree, and native client-agent release notes.
+---
+
+**Released: 2026-08**
+
+NetOpsBench v0.2.0 extends the live benchmark from the original CLOS profiles to Xlarge CLOS and two multi-tier Fat-tree profiles. It also replaces the client-side Python Pingmesh and iperf process fleet with a native Rust client agent while preserving the public observation, tool, evaluator, and SDK contracts.
+
+## Large topology profiles
+
+| Profile | Family | Switches | Clients | Cases | Operational | Agent-scored |
+|---|---|---:|---:|---:|---:|---:|
+| Xlarge | CLOS | 144 | 128 | 70 | 70 / 70 | 70 / 70 |
+| Fat-tree K=8 | Fat-tree | 80 | 128 | 70 | 70 / 70 | 70 / 70 |
+| Fat-tree K=12 | Fat-tree | 180 | 144 | 70 | 70 / 70 | 70 / 70 |
+
+Each profile contains four healthy cases and fault cases spanning physical links, devices, impairments, routing, BGP, policy, and ACL behavior. Fault placement covers access and upper-fabric tiers; endpoint Pingmesh can remain healthy when ECMP successfully preserves forwarding, while interface and control-plane telemetry still expose the degraded redundancy.
+
+## Runtime changes
+
+- A single Rust binary runs independent Pingmesh and background-traffic processes in every client.
+- Pingmesh uses compact topology configuration, kernel receive timestamps, topology-local Telegraf ingestion, and full-window anomaly semantics.
+- Background traffic is runtime-owned and health-checked as a complete flow matrix instead of being recreated for every case.
+- Fault injection and recovery use transactional state tracking, readback, convergence checks, quarantine, and exact resource cleanup.
+- Large BGP collections preserve fast transition/index data while writing stable full snapshots at a lower cadence.
+
+## Compatibility and results
+
+The stable `netopsbench.sdk` agent, scenario, report, and tool entrypoints remain available. The v0.2 result snapshot is a release-validation reference, not a permanent model ranking; provider behavior and agent strategies can change independently of the benchmark contract.
+
+See [Benchmark Results](/docs/run-benchmarks/results) for accuracy, localization, observability, cost, and trace inventory, the [public v0.2 trace release](https://huggingface.co/datasets/yyyyyt/netopsbench-trace/tree/main/releases/netopsbench-0.2) for all 319 ATIF trajectories, and [Benchmark Methodology](/docs/run-benchmarks/methodology) for the exact scoring and detector semantics.
@@ -29,9 +29,9 @@ Pingmesh endpoint density.
 | Small | CLOS | 2 spine + 4 leaf | 8 |
 | Medium | CLOS | 4 spine + 8 leaf | 16 |
 | Large | CLOS | 4 spine + 16 leaf | 64 |
-| XLarge | CLOS | 16 spine + 128 leaf | 128 |
-| Fat-tree K8 | Fat-tree | 16 core + 32 aggregation + 32 edge | 128 |
-| Fat-tree K12 | Fat-tree | 36 core + 72 aggregation + 72 edge | 144 |
+| Xlarge | CLOS | 16 spine + 128 leaf | 128 |
+| Fat-tree K=8 | Fat-tree | 16 core + 32 aggregation + 32 edge | 128 |
+| Fat-tree K=12 | Fat-tree | 36 core + 72 aggregation + 72 edge | 144 |
 
 The benchmark covers twelve canonical fault types:
 
@@ -51,9 +51,9 @@ Each scale combines fault cases with healthy negative samples:
 | Small | 12 | 3 | 15 |
 | Medium | 24 | 4 | 28 |
 | Large | 48 | 4 | 52 |
-| XLarge | 66 | 4 | 70 |
-| Fat-tree K8 | 66 | 4 | 70 |
-| Fat-tree K12 | 66 | 4 | 70 |
+| Xlarge | 66 | 4 | 70 |
+| Fat-tree K=8 | 66 | 4 | 70 |
+| Fat-tree K=12 | 66 | 4 | 70 |
 
 Scenarios are generated from the packaged canonical campaign spec. Running `netopsbench benchmark prepare --scales <scale>` materializes YAML files under `scenarios/generated/<scale>/`; `--spec` can select a custom campaign.
 
@@ -153,3 +153,5 @@ The judge affects only fault-type classification. Verdict, device localization,
 | `avg_tool_calls` | Mean tool calls per case. |
 
 The scoring implementation lives in `netopsbench.evaluator.scorer` and session scoring helpers under `netopsbench.platform.session.scoring`.
+
+The large profiles and native client-agent lifecycle are documented in the [NetOpsBench v0.2.0 release notes](/docs/releases/v0.2.0).
-Original file line number
+Diff line change
@@ @@ -0,0 +1,4 @@ @@
 +{
 +  "title": "Releases",
 +  "pages": ["v0.2.0"]
 +}