docs/qa/feature-user-story-status.csv

feature_id,surface,feature,entrypoint_or_route,source_evidence,user_story,expected_behavior,setup_preconditions,test_command_or_steps,status,first_test_result,error_id,error_summary,fix_status,retest_result,last_verified_at,notes
CLI-001,CLI,Generic model run harness,ctx,pyproject.toml project.scripts,As a user I can run or resume a model-agnostic ctx session,ctx creates or resumes a session with configured provider tools budgets and transcript ordering intact,Provider credentials or dry-run/local test provider available,ctx --help then focused run/resume smoke,Tested Pass,test_harness_cli_run.py: 67 passed for ctx run/resume/sessions.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-002,CLI,Repository recommendation,ctx-recommend,pyproject.toml project.scripts,As a user I can ask ctx for current repo recommendations,ctx returns capped relevant skill agent MCP recommendations from the shared graph engine,Graph installed and target repo available,ctx-recommend --help then sample repo recommendation,Tested Pass,recommend CLI/golden subset: 24 passed for ctx-recommend and shared recommendation surfaces.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-003,CLI,Initial setup,ctx-init,pyproject.toml project.scripts,As a new user I can install hooks graph and model mode from one command,ctx-init installs requested pieces idempotently and reports failures with non-zero exit,Writable user config directory,ctx-init --model-mode skip --graph,Retested Pass,"PASS: temp USERPROFILE ctx-init --model-mode skip --knowledge-mode skip created .claude config/toolboxes and exited 0. PASS: temp USERPROFILE ctx-init --graph --graph-install-mode runtime --model-mode skip installed graph.json and harness page in 129.2s.",CTX-AZ-006,Full graph install exceeded the 10-minute Windows smoke budget before packed wiki artifacts: native-tar full install completed in 759.0s and cleanup took 171.7s.,Packed full artifact now ships wiki-packs and omits expanded non-root markdown fanout; ctx-init checksum updated.,PASS: temp USERPROFILE ctx-init --graph --graph-install-mode full --model-mode skip --knowledge-mode skip exited 0 in 409.7s with graph-store.sqlite3 wiki-packs harness pages and no expanded skills/agents/MCP/converted/concepts markdown.,2026-06-23,Runtime path remains usable; full path now installs packed markdown wiki under the smoke budget while keeping harness pages direct.
CLI-004,CLI,Claude hook install,ctx-install-hooks,pyproject.toml project.scripts,As a Claude Code user I can install ctx hooks,Hooks are installed idempotently and point to existing modules,Claude config path writable,ctx-install-hooks --help then temp-home smoke,Tested Pass,inject hooks/ctx-init hook subset: 24 passed for hook install and failure handling.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-005,CLI,Repo stack scan,ctx-scan-repo,pyproject.toml project.scripts,As a developer I can scan a repo and get stack plus recommendations,Scan detects stack signatures and renders recommendations consistently with graph engine,Sample repos available,ctx-scan-repo --repo . --recommend,Retested Pass,FAIL: first run classified ctx as ml-project because optional embeddings extra contains torch,CTX-AZ-002,Optional pyproject extras were treated as primary project-type dependencies,Fixed by excluding optional/dev pyproject deps from primary project-type classification while keeping them as searchable signals,PASS: focused scan tests 5 passed; live scan now reports Type infrastructure and recommendations render,2026-06-23,
CLI-006,CLI,Skill add,ctx-skill-add,pyproject.toml project.scripts,As a user I can add or update a skill,Skill is validated micro-skill gated written to wiki and attached to graph/index without duplicate stale edges,Temp skill directory and wiki path,ctx-skill-add --skill-path temp/SKILL.md --name demo --wiki temp/wiki,Retested Pass,Failed in isolated A-Z flow: Windows UTF-8 BOM input was rejected by markdown/JSON parsing; some rejected batch CLIs could still exit 0.,CTX-AZ-004,"Entity onboarding was not Windows-BOM-safe across skill/agent/MCP/harness inputs, and failed batches needed non-zero exits.",Fixed,Skill add accepted BOM-prefixed SKILL.md and wrote wiki entity in temp CLI flow. Focused pytest regressions passed.,2026-06-23,Regression coverage added for BOM inputs and MCP rejected-batch exit.
CLI-007,CLI,Skill install,ctx-skill-install,pyproject.toml project.scripts,As a Claude Code user I can install a recommended skill,Skill installs into Claude skills directory with manifest and safe overwrite behavior,Temp Claude home,ctx-skill-install --help then dry/smoke path if available,Tested Pass,"install/unload batch: 187 passed, including skill install behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-008,CLI,Skill unload,ctx-skill-unload,pyproject.toml project.scripts,As a user I can unload stale skills,Skill is disabled/removed from active load state without deleting unrelated user files,Temp installed skill,ctx-skill-unload --help then temp uninstall smoke,Tested Pass,"install/unload batch: 187 passed, including skill unload behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-009,CLI,Skill mirror,ctx-skill-mirror,pyproject.toml project.scripts,As a user I can mirror installed skills into wiki,Mirror preserves source metadata and does not overwrite higher quality records silently,Temp skill source,ctx-skill-mirror --help,Tested Pass,"mirror/quality batch: 123 passed, including skill mirror behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-010,CLI,Skill health,ctx-skill-health,pyproject.toml project.scripts,As a maintainer I can check skill health,Health check reports missing frontmatter orphan entries and line drift with actionable exits,Repo graph/wiki available,ctx-skill-health --help then focused check,Tested Pass,"SkillSpector/health batch: 56 passed, including skill health behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-011,CLI,Skill quality,ctx-skill-quality,pyproject.toml project.scripts,As a maintainer I can compute skill quality,Quality scores explain telemetry intake graph and routing signals,Repo graph/wiki available,ctx-skill-quality --help then sample,Tested Pass,"mirror/quality batch: 123 passed, including skill quality behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-012,CLI,Agent add,ctx-agent-add,pyproject.toml project.scripts,As a user I can add or update an agent,Agent page is validated written atomically and connected to graph,Temp agent markdown and wiki path,ctx-agent-add --agent-path temp/agent.md --name demo --wiki temp/wiki,Retested Pass,Failed in isolated A-Z flow: Windows UTF-8 BOM input was rejected by markdown/JSON parsing; some rejected batch CLIs could still exit 0.,CTX-AZ-004,"Entity onboarding was not Windows-BOM-safe across skill/agent/MCP/harness inputs, and failed batches needed non-zero exits.",Fixed,Agent add accepted BOM-prefixed agent markdown and wrote wiki entity in temp CLI flow. Focused pytest regressions passed.,2026-06-23,Regression coverage added for BOM inputs and MCP rejected-batch exit.
CLI-013,CLI,Agent install,ctx-agent-install,pyproject.toml project.scripts,As a Claude Code user I can install an agent,Agent installs safely and appears in manifest/load surface,Temp Claude home,ctx-agent-install --help,Tested Pass,"install/unload batch: 187 passed, including agent install behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-014,CLI,Agent unload,ctx-agent-unload,pyproject.toml project.scripts,As a user I can unload stale agents,Agent active state is removed safely and recommendation can suggest replacements,Temp installed agent,ctx-agent-unload --help,Tested Pass,install/unload batch plus monitor 3-type coverage: 187 passed for install/unload core behavior.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-015,CLI,Agent mirror,ctx-agent-mirror,pyproject.toml project.scripts,As a user I can mirror agents into wiki,Mirror emits normalized wiki pages with source metadata,Temp agent source,ctx-agent-mirror --help,Tested Pass,Temp CLI smoke passed: ctx-agent-mirror mirrored agent body into converted-agents and cleaned temp files.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-016,CLI,MCP add,ctx-mcp-add,pyproject.toml project.scripts,As a user I can add or update an MCP server,MCP record is validated enriched written atomically and linked in graph,Temp MCP JSON/URL and wiki path,ctx-mcp-add --dry-run with fixture,Retested Pass,Failed in isolated A-Z flow: Windows UTF-8 BOM input was rejected by markdown/JSON parsing; some rejected batch CLIs could still exit 0.,CTX-AZ-004,"Entity onboarding was not Windows-BOM-safe across skill/agent/MCP/harness inputs, and failed batches needed non-zero exits.",Fixed,MCP add accepted BOM-prefixed JSON and wrote sharded wiki entity in temp CLI flow; stdin BOM regression is covered by pytest. Focused pytest regressions passed.,2026-06-23,Regression coverage added for BOM inputs and MCP rejected-batch exit.
CLI-017,CLI,MCP install,ctx-mcp-install,pyproject.toml project.scripts,As a Claude Code user I can install an MCP server,MCP config is updated safely and secrets are not echoed or persisted inline,Temp Claude config,ctx-mcp-install --help then dry/temp smoke,Tested Pass,"install/unload batch: 187 passed, including MCP install behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-018,CLI,MCP uninstall,ctx-mcp-uninstall,pyproject.toml project.scripts,As a user I can uninstall an MCP server,MCP config entry is removed without damaging other servers,Temp Claude config,ctx-mcp-uninstall --help,Tested Pass,"install/unload batch: 187 passed, including MCP uninstall behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-019,CLI,MCP fetch,ctx-mcp-fetch,pyproject.toml project.scripts,As a maintainer I can fetch MCP catalog data,Fetch produces normalized records with source provenance and retry-safe output,Network available or fixture mode,ctx-mcp-fetch --help,Tested Pass,"MCP fetch/ingest/server batch: 77 passed, 1 skipped.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-020,CLI,MCP ingest,ctx-mcp-ingest,pyproject.toml project.scripts,As a maintainer I can ingest MCP catalog data,Ingest deduplicates normalizes and updates wiki/graph inputs,Fixture catalog,ctx-mcp-ingest --help,Tested Pass,"MCP fetch/ingest/server batch: 77 passed, 1 skipped.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-021,CLI,MCP enrich,ctx-mcp-enrich,pyproject.toml project.scripts,As a maintainer I can enrich MCP records,Enrichment updates metadata without clobbering curated fields,Fixture record,ctx-mcp-enrich --help,Tested Pass,"MCP enrich/canonical/quality signals batch: 78 passed, 1 skipped.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-022,CLI,MCP quality,ctx-mcp-quality,pyproject.toml project.scripts,As a maintainer I can score MCP quality,MCP score explains source docs install and metadata signals,Repo wiki available,ctx-mcp-quality --help,Tested Pass,"MCP quality batches passed: 123 passed plus 78 passed, 1 skipped in signal/enrich coverage.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-023,CLI,MCP rebuild index,ctx-mcp-rebuild-index,pyproject.toml project.scripts,As a maintainer I can rebuild MCP lookup index,Index rebuild is deterministic and validates duplicate slugs,Repo wiki available,ctx-mcp-rebuild-index --help,Tested Pass,Temp CLI smoke passed: ctx-mcp-rebuild-index wrote .canonical-index.json; canonical index tests included in 78 passed batch.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-024,CLI,MCP server,ctx-mcp-server,pyproject.toml project.scripts,As an MCP client I can connect to ctx as an MCP server,Server exposes recommendation/query tools and handles invalid requests safely,Local MCP client or stdio smoke,ctx-mcp-server --help,Tested Pass,"MCP fetch/ingest/server batch: 77 passed, 1 skipped.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-025,CLI,Harness add,ctx-harness-add,pyproject.toml project.scripts,As a user I can add or update a harness,Harness page is atomic symlink-safe scored and linked to compatible model goals,Temp harness JSON/wiki,ctx-harness-add --dry-run --from-json fixture,Retested Pass,Failed in isolated A-Z flow: Windows UTF-8 BOM input was rejected by markdown/JSON parsing; some rejected batch CLIs could still exit 0.,CTX-AZ-004,"Entity onboarding was not Windows-BOM-safe across skill/agent/MCP/harness inputs, and failed batches needed non-zero exits.",Fixed,Harness add accepted BOM-prefixed JSON and wrote wiki entity in temp CLI flow. Focused pytest regressions passed.,2026-06-23,Regression coverage added for BOM inputs and MCP rejected-batch exit.
CLI-026,CLI,Harness install/recommend,ctx-harness-install,pyproject.toml project.scripts,As a local/API model user I can get harness recommendations and install one,Wizard interviews model goal tools privacy runtime and recommends harness above threshold or no-fit PRD,Graph/wiki installed,ctx-harness-install --dry-run --goal sample --model openai/gpt-4.1,Retested Pass,FAIL: generic Python coding-agent goal recommended narrow text-to-cad harness,CTX-AZ-003,Narrow-domain harnesses could win generic goals via shared python/filesystem/verification terms,Fixed with domain-fit penalty unless the user goal names the harness domain,PASS: generic goal now returns no-fit PRD; explicit CAD/robotics goal still recommends text-to-cad; focused tests 4 passed,2026-06-23,
CLI-027,CLI,Bundle suggest,ctx-bundle-suggest,pyproject.toml project.scripts,As a session-start hook I can get a capped bundle suggestion,Bundle suggestions are capped deduped and explain why each entity is useful,Graph installed and repo context,ctx-bundle-suggest --help,Tested Pass,"bundle/lifecycle/toolbox batch: 110 passed, including bundle suggestion/orchestration behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-028,CLI,Lifecycle,ctx-lifecycle,pyproject.toml project.scripts,As a user I can track load/unload lifecycle,Usage and stale entities are tracked with confirmable unload behavior,Audit/log path available,ctx-lifecycle --help,Tested Pass,"bundle/lifecycle/toolbox batch: 110 passed, including lifecycle behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-029,CLI,Toolbox,ctx-toolbox,pyproject.toml project.scripts,As a user I can manage toolbox councils,Toolbox init/list/run is idempotent and does not warn on expected existing state,Temp toolbox home,ctx-toolbox --help,Tested Pass,"bundle/lifecycle/toolbox batch: 110 passed, including toolbox config/hooks behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-030,CLI,Dedup check,ctx-dedup-check,pyproject.toml project.scripts,As a maintainer I can block duplicate entities,Dedup gate catches near duplicates and reports merge/update advice,Repo wiki/graph available,ctx-dedup-check --help then sample,Tested Pass,"wiki queue/dedup/source/package/category batch: 106 passed, including dedup behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-031,CLI,Tag backfill,ctx-tag-backfill,pyproject.toml project.scripts,As a maintainer I can backfill weak tags,Backfill adds useful tags without overwriting curated tags blindly,Fixture wiki,ctx-tag-backfill --help,Tested Pass,Temp-home CLI smoke passed: ctx-tag-backfill generated a report for an empty-tag skill; generated report artifact was removed.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-032,CLI,Source registry,ctx-source-registry,pyproject.toml project.scripts,As a maintainer I can track external sources,Registry records source status license and update provenance,Repo source registry available,ctx-source-registry --help,Tested Pass,Source registry test plus CLI JSON smoke passed; source registry emitted validated JSON and temp output was removed.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-033,CLI,Wiki graphify,ctx-wiki-graphify,pyproject.toml project.scripts,As a maintainer I can build the knowledge graph,Graphify exports crash-safe graph artifacts refreshes generated links and validates counts,Repo wiki path,ctx-wiki-graphify --dry-run or graph-only fixture,Tested Pass,wiki graphify/security/density/quality/harness entity tests: 43 passed.,,,,,2026-06-23,Validated in final A-Z pass.
CLI-034,CLI,Incremental attach,ctx-incremental-attach,pyproject.toml project.scripts,As a maintainer I can attach new nodes fast,Attach uses pinned embedding/index geometry writes provenance-rich overlays and supports dry-run/queued fallback,Vector index and fixture node,ctx-incremental-attach attach --dry-run ...,Tested Pass,"graph packs/store/incremental batch: 59 passed, including incremental attach A-Z flow.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-035,CLI,Incremental shadow,ctx-incremental-shadow,pyproject.toml project.scripts,As a maintainer I can compare incremental attach with batch graph,Shadow mode reports precision recall top-k agreement and bad examples,Graph/index available,ctx-incremental-shadow --help,Tested Pass,incremental shadow tests: 6 passed.,,,,,2026-06-23,Validated in final A-Z pass.
CLI-036,CLI,Graph store,ctx-graph-store,pyproject.toml project.scripts,As a dashboard user I can search graph quickly via SQLite store,Store builds validates searches and returns neighborhoods consistent with graph artifacts,Graph artifact available,ctx-graph-store --help,Tested Pass,"graph packs/store/incremental batch: 59 passed, including graph-store CLI build/validate/search/neighborhood behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-037,CLI,Pack compaction,ctx-pack-compact,pyproject.toml project.scripts,As a maintainer I can compact overlay packs,Compaction validates overlays tombstones counts and writes new base artifacts atomically,Graph packs available,ctx-pack-compact --help,Tested Pass,graph packs/store/incremental batch: 59 passed plus pack_compaction coverage.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-038,CLI,Wiki worker,ctx-wiki-worker,pyproject.toml project.scripts,As a maintainer I can drain wiki maintenance queue,Worker leases retries jobs and recovers after failure without corrupting artifacts,Temp wiki queue,ctx-wiki-worker --once --limit 1 on fixture,Tested Pass,"wiki queue/dedup/source/package/category batch: 106 passed, including wiki worker upsert/delete/compaction jobs.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-039,CLI,SkillSpector audit,ctx-skillspector-audit,pyproject.toml project.scripts,As a maintainer I can audit skill quality and stale resources,Audit reports stale wrappers invalid references and remediation actions,Repo skills available,ctx-skillspector-audit --help,Tested Pass,"SkillSpector/health batch: 56 passed, including audit payload/report normalization.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-040,CLI,SkillSpector scan,ctx-skillspector-scan,pyproject.toml project.scripts,As a maintainer I can scan SkillSpector data for dashboard,Scan produces dashboard-readable results without blocking normal monitor load,Repo skills available,ctx-skillspector-scan --help,Tested Pass,"SkillSpector/health batch: 56 passed, including scan/service behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-041,CLI,SkillSpector remediation,ctx-skillspector-remediation,pyproject.toml project.scripts,As a maintainer I can remediate SkillSpector findings,Remediation proposes or applies safe fixes with audit evidence,Audit fixture,ctx-skillspector-remediation --help,Tested Pass,"SkillSpector/health batch: 56 passed, including remediation behavior.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
CLI-042,Dashboard,Monitor server,ctx-monitor,pyproject.toml project.scripts,As a user I can launch the local dashboard,Dashboard starts on local host serves all nav pages and does not expose LAN mode unauthenticated,Graph runtime installed,ctx-monitor serve --host 127.0.0.1 --port 8765,Tested Pass,Started ctx-monitor on 127.0.0.1:8765; HTTP dashboard responded after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-001,Dashboard,Home page,/,src/ctx/monitor/routes.py,As a user I can land on a useful dashboard home,Home shows current counts and high-value entry points without stale numbers,Monitor running,GET /,Tested Pass,GET / returned 200 in 1506ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-002,Dashboard,Loaded page,/loaded,src/ctx/monitor/routes.py,As a user I can see loaded entities,Loaded page lists active skills agents MCPs and supports load state clarity,Monitor running,GET /loaded,Tested Pass,GET /loaded returned 200 in 79ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-003,Dashboard,Skills page,/skills,src/ctx/monitor/routes.py,As a user I can browse skills with pagination/search,Skills page is fast and does not dump all cards at once,Monitor running,GET /skills,Tested Pass,GET /skills returned 200 in 165ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-004,Dashboard,SkillSpector page,/skillspector,src/ctx/monitor/routes.py,As a maintainer I can inspect quality findings,SkillSpector page renders findings and remediation links,Monitor running,GET /skillspector,Tested Pass,GET /skillspector returned 200 in 3315ms after restart; functional but slower than core pages.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-005,Dashboard,Wiki index/catalog,"/wiki, /catalog, and /catalog/",src/ctx/monitor/routes.py,As a user I can search entity catalog,Wiki/catalog supports autocomplete-like search and tiles for skill agent MCP harness,Monitor running,GET /catalog?type=agent&q=review,Retested Pass,FAIL: live route timed out after 60018ms before fix,CTX-AZ-001,/catalog with query rebuilt too much server-side catalog state while /api/entities/search was fast,Fixed by using bounded search results for queried wiki/catalog render,PASS: live route 200 in 1422ms after monitor restart; focused tests 4 passed,2026-06-23,Regression covered by test_render_wiki_index_query_uses_bounded_search
DASH-006,Dashboard,Wiki entity page,/wiki/<slug>,src/ctx/monitor/routes.py,As a user I can inspect a full entity page,Entity page has Overview Subgraph Quality tabs and rendered markdown/frontmatter,Monitor running,GET /wiki/github?type=mcp-server,Tested Pass,GET /wiki/github?type=mcp-server returned 200 in 2705ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-007,Dashboard,Graph page,/graph,src/ctx/monitor/routes.py,As a user I can explore graph interactively,Graph loads live search results click-to-focus double-click-back filters match histogram and useful inspector,Monitor running,GET /graph then interact,Retested Pass,PASS: /graph returned 200 in 31ms after monitor restart,,,,"PASS: Playwright test_ctx_monitor_browser.py::test_graph_page_uses_builtin_svg_renderer covered renderer click focus double-click back filters histogram live search type shapes and inspector.",2026-06-23,Browser interaction pass completed.
DASH-008,Dashboard,Manage page,/manage,src/ctx/monitor/routes.py,As a user I can manage entities,Manage supports search upload/upsert delete/manual edit with safety prompts,Monitor running,GET /manage,Tested Pass,GET /manage returned 200 in 261ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-009,Dashboard,Harness setup,/harness,src/ctx/monitor/routes.py,As a local/API model user I can use a harness wizard,Harness Setup interviews model and goals recommends matching harness or no-fit PRD,Monitor running,GET /harness,Tested Pass,GET /harness returned 200 in 23ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-010,Dashboard,Docs page,/docs,src/ctx/monitor/routes.py,As a user I can read repo docs inside dashboard,Docs render markdown like mkdocs with tabs sidebar hierarchy and local search,Monitor running,GET /docs,Tested Pass,GET /docs returned 200 in 1219ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-011,Dashboard,Config page,/config,src/ctx/monitor/routes.py,As a user I can view/edit important config,Config shows explanations defaults required fields validation and safe save,Monitor running,GET /config and POST /api/config,Tested Pass,GET /config returned 200 in 12ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-012,Dashboard,Status page,/status,src/ctx/monitor/routes.py,As a user I can see system health,Status reports graph/wiki/cache/session health with actionable errors,Monitor running,GET /status,Tested Pass,GET /status returned 200 in 3750ms after restart; functional but slower than core pages.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-013,Dashboard,KPI page,/kpi,src/ctx/monitor/routes.py,As a user I can inspect recommendation effectiveness,KPIs show usage score quality and lifecycle trends without slow unbounded render,Monitor running,GET /kpi,Tested Pass,GET /kpi returned 200 in 107ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-014,Dashboard,Runtime page,/runtime,src/ctx/monitor/routes.py,As a user I can inspect runtime artifacts,Runtime shows graph store cache versions and artifact freshness,Monitor running,GET /runtime,Tested Pass,GET /runtime returned 200 in 25ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-015,Dashboard,Sessions page,/sessions,src/ctx/monitor/routes.py,As a user I can inspect sessions,Sessions page lists transcripts and details without leaking secrets,Monitor running,GET /sessions,Tested Pass,GET /sessions returned 200 in 269ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-016,Dashboard,Logs page,/logs,src/ctx/monitor/routes.py,As a user I can inspect logs,Logs render audit/log files or clear empty state,Monitor running,GET /logs,Tested Pass,GET /logs returned 200 in 132ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-017,Dashboard,Live events,/events and /live,src/ctx/monitor/routes.py,As a user I can tail live events,Live page connects SSE and shows clear empty/disconnected states,Monitor running,GET /events and /api/events.stream,Tested Pass,GET /live returned 200 in 87ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
DASH-018,Dashboard,Session detail page,/session/<session_id>,src/ctx/monitor/routes.py,As a user I can inspect one session timeline,Session detail filters audit and load events to the requested session and escapes untrusted values,Monitor running with audit fixtures,GET /session/<session_id>,Tested Pass,Focused monitor tests cover session_detail filtering and render_session_detail escaping.,,,,,2026-06-23,Added during route inventory coverage audit.
DASH-019,Dashboard,Sidecar detail page,/skill/<slug>,src/ctx/monitor/routes.py,As a user I can inspect one entity sidecar,Sidecar detail renders quality breakdown and timeline for typed skill agent MCP or harness sidecars,Monitor running with sidecar fixtures,GET /skill/<slug>?type=agent,Tested Pass,Focused monitor tests cover sidecar links plus render_skill_detail route wiring and typed sidecar loading.,,,,,2026-06-23,Added during route inventory coverage audit.
API-001,Dashboard API,Sessions API,/api/sessions.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch sessions JSON,Endpoint returns valid JSON quickly with no secrets,Monitor running,GET /api/sessions.json,Tested Pass,GET /api/sessions.json returned 200 in 241ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
API-002,Dashboard API,Manifest API,/api/manifest.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch manifest JSON,Endpoint returns loaded entities and counts consistently,Monitor running,GET /api/manifest.json,Tested Pass,GET /api/manifest.json returned 200 in 5ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
API-003,Dashboard API,Status API,/api/status.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch health JSON,Endpoint returns graph/wiki/cache health with correct status code,Monitor running,GET /api/status.json,Tested Pass,GET /api/status.json returned 200 in 4303ms after restart; functional but slower than core APIs.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
API-004,Dashboard API,KPI API,/api/kpi.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch KPI JSON,Endpoint returns bounded KPI payload quickly,Monitor running,GET /api/kpi.json,Tested Pass,GET /api/kpi.json returned 200 in 105ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
API-005,Dashboard API,Grades API,/api/grades.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch grades JSON,Endpoint returns score and grade payload with explanations,Monitor running,GET /api/grades.json,Retested Pass,Initial dashboard sweep: GET /api/grades.json returned 200 but took 16833ms for a tiny payload.,CTX-AZ-005,Grades API ignored cached KPI grade_counts and rescanned all quality sidecars on demand.,Fixed,"After routing through cached KPI summary, GET /api/grades.json returned 200 in 92ms after restart.",2026-06-23,Fallback full sidecar scan remains for missing KPI summary.
API-006,Dashboard API,Runtime API,/api/runtime.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch runtime JSON,Endpoint returns artifact/cache versions and paths safely,Monitor running,GET /api/runtime.json,Tested Pass,GET /api/runtime.json returned 200 in 21ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
API-007,Dashboard API,Entity search API,/api/entities/search.json,src/ctx/monitor/routes.py,As a dashboard client I can search entities,Endpoint supports live query type filters and useful fuzzy results,Monitor running,GET /api/entities/search.json?q=review&type=agent,Tested Pass,PASS: endpoint returned 200 in 28ms with agent review results,,,,,2026-06-23,
API-008,Dashboard API,Entity detail API,/api/entity/<slug>.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch entity details,Endpoint returns canonical entity docs frontmatter and graph sidecar,Monitor running,GET /api/entity/github.json?type=mcp-server,Tested Pass,GET /api/entity/github.json?type=mcp-server returned 200 in 27ms after restart.,,,,,2026-06-23,Dashboard route sweep; no mutation performed.
API-009,Dashboard API,Graph neighborhood API,/api/graph/<slug>.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch a graph neighborhood,Endpoint returns bounded nodes edges scores and reasons fast after runtime install,Monitor running,GET /api/graph/github.json?type=skill&limit=20,Tested Pass,PASS: endpoint returned 200 in 2295ms with nodes/edges JSON,,,,,2026-06-23,
API-010,Dashboard API,Load/unload APIs,/api/load and /api/unload,src/ctx/monitor/routes.py,As a dashboard user I can load/unload entities,POST validates input changes active state and rejects unsafe requests,Monitor running and temp state,POST /api/load then /api/unload,Tested Pass,"Focused pytest passed: perform_load rejects invalid slugs, load security scan surfaces output, perform_unload rejects invalid slugs.",,,,,2026-06-23,"Validated through isolated monitor mutation tests, not live local catalog mutation."
API-011,Dashboard API,Entity upsert/delete APIs,/api/entity/upsert and /api/entity/delete,src/ctx/monitor/routes.py,As a dashboard user I can manually add edit or delete entities,POST validates type slug content and queues graph/wiki updates safely,Monitor running and temp wiki,POST fixture upsert/delete,Tested Pass,Focused pytest passed: entity upsert writes wiki page and queues graph refresh; existing-page confirmation and delete behavior covered.,,,,,2026-06-23,"Validated through isolated monitor mutation tests, not live local catalog mutation."
API-012,Dashboard API,Config read API,/api/config.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch effective config JSON,Endpoint returns defaults plus user overrides with required config metadata and no secrets,Monitor running,GET /api/config.json,Tested Pass,Focused monitor config tests cover effective payload rendering and config save/clear behavior.,,,,,2026-06-23,Added during route inventory coverage audit.
API-013,Dashboard API,Sidecars API,/api/sidecars.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch paginated sidecar cards,Endpoint supports query type grade floor and limit filters without full unbounded scans,Monitor running with sidecar fixtures,GET /api/sidecars.json?q=review&type=agent&grade=A&hide_floor=1&limit=1,Tested Pass,Focused monitor API test returned filtered sidecar JSON with total limit filters and caps.,,,,,2026-06-23,Added during route inventory coverage audit.
API-014,Dashboard API,SkillSpector API,/api/skillspector.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch SkillSpector findings,Endpoint returns filtered audit summary and records for dashboard remediation views,Monitor running with skillspector audit fixture,GET /api/skillspector.json?severity=HIGH&q=prompt,Tested Pass,Focused monitor API test returned HIGH prompt-injection SkillSpector record JSON.,,,,,2026-06-23,Added during route inventory coverage audit.
API-015,Dashboard API,Sidecar detail API,/api/skill/<slug>.json,src/ctx/monitor/routes.py,As a dashboard client I can fetch one sidecar JSON,Endpoint returns typed sidecar details or a safe 404 for missing/unsafe slugs,Monitor running with typed sidecar fixture,GET /api/skill/filesystem.json?type=mcp-server,Tested Pass,Focused monitor API test returned typed MCP sidecar JSON and unsafe slug rejection is covered.,,,,,2026-06-23,Added during route inventory coverage audit.
GRAPH-001,Graph/Wiki,Full graph artifact validation,src/validate_graph_artifacts.py,pyproject and scripts/ci_preflight.py,As a maintainer I can validate shipped graph artifacts,Validation deep-checks tar members counts no originals no locks exact release counts and line thresholds,Repo graph artifacts present,python src/validate_graph_artifacts.py --deep ...,Tested Pass,PASS: exact counts validated nodes 79958 edges 1778069 semantic 1088763 pages skills 68494 agents 467 MCPs 10790 harnesses 207,,,,,2026-06-23,
GRAPH-002,Graph/Wiki,Published stats sync,src/update_repo_stats.py,pyproject and src/update_repo_stats.py,As a maintainer I can keep README docs and About counts in sync,Stats come from shipped graph artifacts and checks fail when docs drift,Repo graph artifacts present,python src/update_repo_stats.py --check,Tested Pass,PASS: README/docs stats are up to date; test_release_graph_count_sync 3 passed,,,,,2026-06-23,
GRAPH-003,Graph/Wiki,Graph pack manifest,ctx core graph_packs/pack_validation,src/ctx/core/graph and wiki modules,As a maintainer I can use base and overlay graph packs,Packs have manifest export IDs parent IDs checksums tombstones and merged reads,Graph pack fixtures,pytest graph pack tests,Tested Pass,graph packs/store/incremental batch: 59 passed and graph pack manifest tests passed.,,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
GRAPH-004,Graph/Wiki,SQLite operational graph store,ctx-graph-store,src/ctx/core/graph/graph_store.py,As a dashboard user I get fast search/neighborhoods,Store builds from graphify-out validates freshness and accelerates reads,Runtime graph installed,ctx-graph-store build/search/neighborhood fixture,Tested Pass,"graph packs/store/incremental batch: 59 passed, including SQLite graph store tests.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
GRAPH-005,Graph/Wiki,Vector index and ANN attach,ctx-incremental-attach,src/ctx/core/graph/vector_index.py,As a maintainer I can update graph without global rebuild,New node embedding queries base+delta index and writes overlay edges with provenance,Index fixture,incremental attach dry-run and validate-indexes,Tested Pass,"graph packs/store/incremental batch: 59 passed, including vector/incremental attach A-Z tests.",,,,,2026-06-23,Validated in focused pytest or isolated temp CLI smoke during A-Z pass.
HARNESS-001,Harness,Own-model harness recommendation,ctx-init and ctx-harness-install,ctx_init.py and harness_install.py,As a user with non-Claude model I can choose model/goals and get best harness,Flow interviews provider model goal runtime tools privacy and returns top harnesses above configured score,Runtime graph installed,ctx-init --model-mode custom or harness dry-run,Retested Pass,FAIL: generic own-model coding-agent goal returned text-to-cad false positive,CTX-AZ-003,Narrow-domain harnesses were not penalized for generic goals,Fixed with domain-fit penalty preserving explicit domain matches,PASS: generic coding-agent goal returns no recommendations plus custom harness PRD,2026-06-23,
HARNESS-002,Harness,No-fit harness PRD,ctx-harness-install,harness_install.py,As a user with no fitting harness I get instructions to build one,No-fit path produces provider-aware PRD attachable to ctx recommendation system,Graph with threshold forcing no match,ctx-harness-install --dry-run with restrictive constraints,Tested Pass,PASS: --plan-on-no-fit printed provider-aware Custom Harness PRD for openai/gpt-4.1 with runtime/tools/privacy/verification fields,,,,,2026-06-23,
DOC-001,Docs,Public docs install path,docs/index.md and docs/*.md,docs file inventory,As a first-time user I can follow docs exactly,Docs commands are current numbers accurate no dead module paths no future-roadmap fluff,Docs built locally,mkdocs build or docs route smoke,Tested Pass,python -m mkdocs build --strict passed; python src/update_repo_stats.py --check passed.,,,,,2026-06-23,Validated in final A-Z pass.
DOC-002,Docs,README badges/catalog links,README.md and docs/catalog.md,README plus public docs,As a GitHub visitor I can click badges and browse catalog,Skill agent MCP harness badges open public catalog pages with working search guidance,Public docs available,Inspect README links and catalog route,Tested Pass,"README top badges checked for Tests, Graph, Skills, Agents, MCPs, Harnesses; update_repo_stats --check passed.",,,,,2026-06-23,Validated in final A-Z pass.
DIST-001,Distribution,GitHub About metadata,GitHub repository About,src/update_repo_stats.py,As a GitHub visitor I can understand ctx positioning and current counts from the repo About,About says ctx is a recommendation layer not a marketplace and uses shipped graph counts,GitHub CLI authenticated,python src/update_repo_stats.py --check-github-about,Tested Pass,GitHub About description is up to date.,,,,,2026-06-23,Validated after the recommendation-layer positioning update.
DIST-002,Distribution,Main CI test workflow,.github/workflows/test.yml,.github/workflows/test.yml,As a contributor I can trust the required main test workflow,Main push runs static gates unit-linux canaries package smoke browser security and required CI aggregate,GitHub Actions available,gh run list --branch main,Tested Pass,Main Tests workflow for Hydrate Hugging Face sync artifacts by checksum (#195) completed successfully.,,,,,2026-06-23,Observed from GitHub Actions run 28053096394.
DIST-003,Distribution,Docs deploy workflow,.github/workflows/docs.yml,.github/workflows/docs.yml,As a docs visitor I get the current public docs after main updates,Docs workflow builds strict MkDocs and deploys GitHub Pages without stale generated content,GitHub Pages enabled,gh run list --branch main,Tested Pass,Deploy docs to GitHub Pages workflow for Align graph validator defaults with shipped graph (#194) completed successfully.,,,,,2026-06-23,Observed from GitHub Actions run 28052467169; #195 did not touch docs.
DIST-004,Distribution,Hugging Face sync workflow,.github/workflows/huggingface-sync.yml,.github/workflows/huggingface-sync.yml and scripts/sync_huggingface.py,As a Hugging Face visitor I get the same tracked repo and graph artifacts as GitHub,Workflow hydrates graph artifacts by matching LFS sha/size release assets before publishing and refuses stale README/docs stats,HF_TOKEN configured in GitHub secrets,GitHub Actions Sync Hugging Face workflow on main,Retested Pass,FAIL: older workflow hydrated the latest release tarball and update_repo_stats --check saw stale 314 MiB docs value.,CTX-AZ-007,Latest-release artifact hydration could publish stale graph tarball to Hugging Face,Fixed by checksum and size release-asset resolver in PR #195,PASS: main Sync Hugging Face workflow completed successfully after PR #195.,2026-06-23,Observed from GitHub Actions run 28053096390.
DIST-005,Distribution,Clean host contract workflow,.github/workflows/clean-host-contract.yml,.github/workflows/clean-host-contract.yml,As a new installer I can verify ctx works on a clean host,Workflow installs the package in a clean environment and runs the clean-host smoke contract,GitHub Actions available,gh run list --workflow clean-host-contract or required checks,Tested Pass,Clean host contract job passed inside the PR #195 required checks.,,,,,2026-06-23,Observed from PR #195 checks before merge.
DIST-006,Distribution,PyPI publish workflow,.github/workflows/publish.yml,.github/workflows/publish.yml,As a PyPI user I can install the tagged release with matching graph artifacts,Tag publish builds the wheel validates graph artifacts uploads release assets and publishes the package,Next release tag available,Tag next patch release and inspect publish workflow result,Needs Validation,Inventoried from .github/workflows/publish.yml; tag-driven publish was not exercised in this tracker slice.,,,,,2026-06-23,Validate on the next release tag before marking pass.
DIST-007,Distribution,Parallel pytest experiment workflow,.github/workflows/xdist-experiment.yml,.github/workflows/xdist-experiment.yml,As a maintainer I can benchmark pytest-xdist safely before adopting it,Workflow runs only when explicitly dispatched and does not affect required checks until proven stable,GitHub Actions available,Manual workflow_dispatch run for xdist experiment,Tested Pass,GitHub Actions run 28133264595 passed on ubuntu-latest windows-latest and macos-latest with Python 3.12.,,,,,2026-06-24,Dispatched manually on main; optional experiment remains out of required checks.
MAINT-001,Maintainer,Backup snapshot audit,scripts/audit_backup.py,scripts/audit_backup.py,As a maintainer I can inspect backup snapshots for missing expected files,Script summarizes backup contents and flags missing graph/wiki/report artifacts without mutating the repo,Backup fixture available,Run script against representative backup fixture,Retested Pass,FAIL: BOM-prefixed manifest.json from PowerShell Set-Content -Encoding UTF8 raised JSONDecodeError.,CTX-AZ-008,Backup audit read manifest JSON with utf-8 and rejected common Windows BOM input,Fixed by reading manifest.json with utf-8-sig,PASS: temp backup fixture with BOM manifest printed entries and OK coverage; regression pytest passed.,2026-06-23,
MAINT-002,Maintainer,Dashboard graph index builder,scripts/build_dashboard_graph_index.py,scripts/build_dashboard_graph_index.py,As a dashboard user I get fast graph search and neighborhoods,Script builds the dashboard SQLite/index artifacts from graph exports for fast monitor reads,Graph fixture available,Run script against small graph fixture and query resulting index,Retested Pass,FAIL: BOM-prefixed graph.json from PowerShell Set-Content -Encoding UTF8 raised JSONDecodeError.,CTX-AZ-008,Dashboard graph index builder read graph JSON with utf-8 and rejected common Windows BOM input,Fixed by reading graph JSON with utf-8-sig,PASS: temp graph fixture built SQLite index with 2 nodes and neighbor mcp-server:github; regression pytest passed.,2026-06-23,
MAINT-003,Maintainer,CI change classifier,scripts/ci_classifier.py,scripts/ci_classifier.py and .github/workflows/test.yml,As a contributor I get the right CI gates for the files I changed,Classifier marks docs graph source package browser and similarity changes conservatively,CI classifier tests available,python -m pytest src/tests/test_ci_classifier.py -q,Tested Pass,CI classifier suite covers path classes and workflow contract behavior.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
MAINT-004,Maintainer,No-test policy gate,scripts/ci_no_test_policy.py,scripts/ci_no_test_policy.py and .github/workflows/test.yml,As a reviewer I can reject risky code or workflow changes without test updates,Policy fails contract changes that omit matching tests while allowing metadata-only releases,CI policy tests available,python -m pytest src/tests/test_ci_classifier.py -q,Tested Pass,CI classifier suite covers no-test policy contract files and release metadata exceptions.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
MAINT-005,Maintainer,Required CI aggregate,scripts/ci_required.py,scripts/ci_required.py and .github/workflows/test.yml,As a branch maintainer I can trust the required CI summary,Aggregate rejects failed missing or wrongly skipped required jobs and allows documented conditional skips,CI required tests available,python -m pytest src/tests/test_ci_classifier.py -q,Tested Pass,CI classifier suite covers required job dependency success failure and conditional skip cases.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
MAINT-006,Maintainer,CI preflight gate,scripts/ci_preflight.py,scripts/ci_preflight.py and scripts/ci_classifier.py,As a contributor I can run the same preflight logic locally that CI uses,Preflight classifies changes then emits the exact docs graph browser package and similarity gates required,Repo checkout with git available,Run scripts/ci_preflight.py against representative changed files,Tested Pass,"select_checks(base_ref='origin/main', files=['README.md','docs/index.md'], profile='quick') returned whitespace repo stats and docs strict build.",,,,,2026-06-23,Validated through the script's public selection seam without launching expensive full preflight.
MAINT-007,Maintainer,Clean host contract script,scripts/clean_host_contract.py,scripts/clean_host_contract.py and .github/workflows/clean-host-contract.yml,As a new installer I can smoke ctx on a clean host,Script builds a clean venv installs ctx and runs init scan harness and monitor checks,Network and clean temp venv available,GitHub clean-host contract or local script smoke,Tested Pass,Clean host contract job passed inside PR #195 required checks.,,,,,2026-06-23,Workflow evidence covers the script path.
MAINT-008,Maintainer,Dashboard smoke script,scripts/dashboard_smoke.py,scripts/dashboard_smoke.py,As a maintainer I can smoke the monitor routes before release,Script starts or probes ctx-monitor and checks key pages/APIs with optional warmup,Monitor smoke tests available,python -m pytest src/tests/test_dashboard_smoke.py -q,Tested Pass,Dashboard smoke test suite covers command construction route probing and failure reporting.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
MAINT-009,Maintainer,Graph artifact guard,scripts/graph_artifact_guard.py,scripts/graph_artifact_guard.py,As a maintainer I can prevent huge graph artifacts from bloating local Git state,Guard parks unparks and prunes graph artifacts safely without deleting unrelated files,Graph artifact fixture available,python -m pytest src/tests/test_graph_artifact_guard.py -q,Tested Pass,Graph artifact guard suite covers park unpark prune and explicit git prune safety.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
MAINT-010,Maintainer,Overlay wiki entities,scripts/overlay_wiki_entities.py,scripts/overlay_wiki_entities.py,As a graph maintainer I can apply curated entity overlays to the wiki artifact,Script overlays entity pages and rebuilds dashboard index without corrupting packed wiki structure,Overlay fixture available,python -m pytest src/tests/test_overlay_wiki_entities.py -q,Tested Pass,Overlay wiki entities tests cover page generation replacements and CLI help.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
MAINT-011,Maintainer,Full wiki tar repacker,scripts/pack_full_wiki_tar.py,scripts/pack_full_wiki_tar.py,As a maintainer I can ship a packed full LLM-wiki artifact without huge expanded fanout,Repacker moves high-fanout pages into wiki packs omits raw originals and is idempotent,Wiki tar fixture available,python -m pytest src/tests/test_pack_full_wiki_tar.py -q,Tested Pass,Pack full wiki tar tests cover packed wiki output original omission and idempotent repack.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
MAINT-012,Maintainer,SkillSpector wiki pruner,scripts/prune_skillspector_wiki.py,scripts/prune_skillspector_wiki.py,As a maintainer I can prune low-value SkillSpector wiki noise before shipping,Script builds pruned artifacts in dry-run/apply modes while preserving required catalog metadata,Prune fixture available,python -m pytest src/tests/test_prune_skillspector_wiki.py -q,Tested Pass,Prune SkillSpector wiki tests cover dry-run apply and artifact rewrite behavior.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
MAINT-013,Maintainer,Hugging Face sync script,scripts/sync_huggingface.py,scripts/sync_huggingface.py and .github/workflows/huggingface-sync.yml,As a Hugging Face visitor I get a clean repo-card snapshot without ignored local artifacts,Script exports tracked files adds repo-card metadata validates graph artifacts and removes stale remote files,HF token or fake API available,python -m pytest src/tests/test_huggingface_sync.py -q,Tested Pass,Hugging Face sync tests cover metadata wrapping stale remote cleanup and hardened workflow contract.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
MAINT-014,Maintainer,Similarity threshold tuner,scripts/tune_similarity_thresholds.py,scripts/tune_similarity_thresholds.py,As a graph maintainer I can calibrate similarity thresholds against fixtures,Tuner sweeps threshold settings and reports quality metrics without mutating graph artifacts,Similarity fixture corpus available,Run tuner against fixture corpus,Tested Pass,python scripts/tune_similarity_thresholds.py completed and printed near/distinct/adversarial score distributions plus threshold sweep.,,,,,2026-06-23,Validated with current fixture corpus; unauthenticated HF warning did not block local cached model load.
PUBLIC-001,Docs,Public catalog search JavaScript,docs/assets/javascripts/catalog.js,docs/assets/javascripts/catalog.js and docs/catalog.md,As a GitHub Pages visitor I can filter catalog tiles without a local dashboard,JavaScript wires input and checkbox filters and updates visible catalog cards/counts client-side,Docs assets available,python -m pytest src/tests/test_docs_catalog_page.py -q,Tested Pass,Docs catalog page tests verify script inclusion input listener and public non-local links.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.
PUBLIC-002,Docs,Public repo stats refresh JavaScript,docs/assets/javascripts/repo-stats-refresh.js,docs/assets/javascripts/repo-stats-refresh.js and mkdocs.yml,As a docs visitor I see current repo stats when GitHub allows client refresh,JavaScript is included by MkDocs and degrades safely when live repo stats cannot be fetched,Docs assets available,python -m pytest src/tests/test_docs_repo_stats.py -q,Tested Pass,Docs repo stats tests verify MkDocs includes the script and the script has the expected refresh behavior.,,,,,2026-06-23,Validated by focused pytest in this tracker slice.