release: 0.16.0

[TheAmericanMaker]

Release 0.16.0 — the field-test round. Four issues found by re-running the full 7-phase deepseek-harness analysis through the published 0.15.0 binary (#111–#114), fixed in #115–#119. Version bumped in all four pinned surfaces (package.json, package-lock.json, server.json ×2, scaffold-version.yaml); release-metadata tests green (341/341).

Local verification

• npm test: 341 pass / 0 fail
• npm run build: clean
• node scripts/smoke-mcp.mjs --tarball codecartographer-pi-0.16.0.tgz: 9/9 ok

Surface verification (per CONTRIBUTING — actual returns, not checkboxes)

Full round-trip through the MCP SDK stdio client against the packed 0.16.0 tarball (init → validate → complete → status → amend → status → guide), exercising every change in this release:

===== codecarto_init =====
Initialized CodeCartographer workspace at /tmp/claude-1000/-home-jamessesler-Documents-Github-experiments-deepseek-harness--claude-worktrees-deepseek-harness-analysis-9eb6e2/4ee8acbd-29dd-445b-b7c9-fcc4b5452a6e/scratchpad/verify016/target/.codecarto.
Pipeline: architecture-only (workflow/pipeline-architecture-only.yaml)
First phase: architecture
Seeded orchestrator files: CONVENTIONS.md, DECISIONS.md (the driving chat holds the orchestrator duties — see GUIDE.md §Roles).

===== codecarto_validate =====
Validation: PASS
Output: .codecarto/findings/architecture/architecture-map.md
NOTE: 5 declared secondary output(s) not written: .codecarto/findings/public-surfaces/public-surfaces.md, .codecarto/findings/runtime-lifecycle/runtime-lifecycle.md, .codecarto/findings/state-and-storage/state-and-storage.md, .codecarto/findings/build-and-deploy/build-and-deploy.md, .codecarto/findings/config-model/config-model.md — write each, or account for it in Coverage and limits / a routed handoff entry. Non-gating.

===== codecarto_complete =====
Marked architecture complete (validation: PASS).
Next phase: complete
Closeout: .codecarto/closeouts/2026-08-18-architecture.md
Orchestrator checkpoint: 1 decision(s) appended to DECISIONS.md; 1 open question(s) outstanding — re-triage their kind labels before the next phase (GUIDE.md §Roles).
Dashboard refreshed: .codecarto/dashboard.html

===== DECISIONS.md heading check (#111) =====
heading line index: 86; row under it: D001 | Verification decision for the 0.16.0 release round-trip. | 2026-08-18-architecture | closeouts/2026-08-18-architecture.md §Decisions Beyond Prompt (architecture)

===== codecarto_status =====
Phase: complete
Pipeline state: complete
Pipeline: architecture-only (workflow/pipeline-architecture-only.yaml)
Progress: 1/1 complete
Open questions (terminal unresolved): 1
Carry-forward (pipeline phases): 0
Post-pipeline work: 0 pending
Next: All phases complete. Review findings; post-pipeline skills: codecarto_list_skills / codecarto_skill.
      1 open question(s) and 0 post-pipeline item(s) remain — apply resolutions with codecarto_amend (write scratch/amendments/<slug>.yaml from templates/amendment.yaml).
      Dashboard: .codecarto/dashboard.html (refreshed on completion and amendment; codecarto_dashboard re-renders on demand). Usage totals: codecarto_usage.

===== codecarto_amend =====
Amendment applied.
Open questions closed: v-OQ1
Post-pipeline items closed: none
Closeout: .codecarto/closeouts/2026-08-18-amendment-verify.md
Dashboard refreshed: .codecarto/dashboard.html

===== codecarto_status =====
Phase: complete
Pipeline state: complete
Pipeline: architecture-only (workflow/pipeline-architecture-only.yaml)
Progress: 1/1 complete
Open questions (terminal unresolved): 0
Carry-forward (pipeline phases): 0
Post-pipeline work: 0 pending
Next: All phases complete. Review findings; post-pipeline skills: codecarto_list_skills / codecarto_skill.
      Dashboard: .codecarto/dashboard.html (refreshed on completion and amendment; codecarto_dashboard re-renders on demand). Usage totals: codecarto_usage.

===== codecarto_guide topic=library (first 240 chars) =====
# The library: where finished specs go

A CodeCartographer **library** is a directory of published reimplementation-specs with provenance — the bridge between analysis runs and everything downstream. An analysis pipeline ends with a spec in

Reading the evidence against the four issues:

• DECISIONS.md completion-log heading never inserted: substring presence check matches template prose #111: DECISIONS.md heading check — heading line index: 86; row under it: D001 | … — the heading is a real line with the appended row beneath it.
• Dashboard is never generated by the pipeline itself — completion and amend should regenerate it #112: Dashboard refreshed: .codecarto/dashboard.html on both codecarto_complete and codecarto_amend.
• Library/publish loop is unreachable by discovery: no guide topic, no overview mention, no terminal routing #113: codecarto_guide topic=library serves the new topic.
• Terminal completion is a dead end: next_actions never names skills, amend, publish, dashboard, or usage #114 + fix: codecarto_status text renders every next action, not just the first #119: terminal codecarto_status renders the full routing — skills line, 1 open question(s) and 0 post-pipeline item(s) remain — apply resolutions with codecarto_amend… with live counts, dashboard/usage line — and after the amendment closes the question, the amend line drops from both stored state and the text view.

After merge: tag v0.16.0 on the merge commit → CI publishes (npm + MCP Registry + GitHub Release) → verify by re-downloading the published tarball.

🤖 Generated with Claude Code