Skip to content

CLI UX: doctor rationale, Glossary, stability markers (#87, #88, #95)#117

Merged
JRS1986 merged 2 commits into
mainfrom
fix/cli-ux-doctor-glossary-stability
May 25, 2026
Merged

CLI UX: doctor rationale, Glossary, stability markers (#87, #88, #95)#117
JRS1986 merged 2 commits into
mainfrom
fix/cli-ux-doctor-glossary-stability

Conversation

@JRS1986
Copy link
Copy Markdown
Owner

@JRS1986 JRS1986 commented May 25, 2026

Summary

Closes #87, closes #88, closes #95.

Test plan

  • uv run pytest -q — 394 tests pass (was 351); new tests/test_artifacts_and_stability.py covers registries, doctor output rendering, glossary contents, and stability marker invariants.
  • python -m coding_scaffold --help — markers render, footer links to Glossary + Stability.
  • python -m coding_scaffold doctor — every artifact line carries -> <why>.

Closes #87, #88, #95.

- New `artifacts.py` registry: one place that names every scaffold artifact
  doctor/pilot recommend, with a <100-char rationale answering "what does this
  unlock?". Doctor's survey now reads from the registry and renders the rationale
  inline next to each artifact line.
- New `docs/docs/wiki/Glossary.md` covering the terms users see in --help and
  doctor output (artifact, persona, MCP, skill, manifest, scope, maturity, ...).
  Linked from the wiki index, the top-level --help footer, and the doctor output
  footer so new users don't need to leave the CLI to interpret it.
- New `cli_stability.py` registry tagging every top-level command as
  stable/preview/experimental. Markers render in --help next to each command name.
  Wiki `Stability.md` defines the contract per level.
- Tests assert: every doctor artifact has a rationale, the Glossary covers the
  required terms, every visible --help command carries a registry entry, every
  marker is one of the three known levels.
JRS1986 added a commit that referenced this pull request May 25, 2026
Closes #90, closes #91.

- New `personas.py` registry: four personas (beginner, control, security,
  team-lead) with focus area, prioritized artifact keys, and tailored next
  commands. Both `doctor --persona <name>` and `pilot --persona <name>` read
  from the same registry so the CLI cannot drift from the Team-Rollout wiki.
- `doctor --persona security` reorders the artifact survey so policy / MCP /
  permissions appear first, switches the recommendation list to a focused
  policy/MCP recipe, and tightens the ignore-for-now to advanced surfaces
  that genuinely don't matter for a security review.
- New `coding-scaffold tour` command: five-screen, stateless walkthrough
  (what the tool does -> artifact families -> doctor/pilot/setup loop ->
  daily session/eval/team workflow -> where to go next). No files written,
  no commands executed. Mentioned in the README "30-Second Start" as step 0.
- Persona import-time validation: a typo in `Persona.artifact_keys` fails the
  module load instead of producing a silent KeyError at runtime.

Stacked on top of #117 (the artifact registry is the persona registry's input).
@JRS1986 JRS1986 merged commit a936f4b into main May 25, 2026
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

1 participant