agentleFS
JetXu-LLMJetXu-LLM/DocMason/skills/canonical/workspace-doctor/SKILL.md

workspace-doctor

Skill152 starsChanged 4 months agoApache-2.0View on GitHub

Diagnose DocMason workspace readiness and report actionable remediation without mutating the repository.

  • Installs packages

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 5 are distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when the task is to inspect the current workspace and explain what is missing or degraded.

    line 8
  2. Distinctive:

    If the agent cannot inspect local files or run commands, stop and explain that the workspace cannot be diagnosed reliably.

    line 16
  3. Very specific:

    If the environment is not ready and .venv is absent, direct the next action to ./scripts/bootstrap-workspace.sh --yes.

    line 22
  4. Very specific:

    treat the launcher as the governed cold-start preflight and bounded liveness gate for bootstrap Python selection, not as an unbounded trust …

    line 23
  5. Very specific:

    treat mixed and degraded toolchain states as repair-needed, not ordinary ask-time ready

    line 26
  6. Very specific:

    If doctor reports a control-plane blocker or pending confirmation, surface that blocker before lower-severity degraded conditions.

    line 27
  7. Very specific:

    If office-renderer is blocked:

    line 31
  8. Very specific:

    If the normal launcher or prepare path still cannot complete honestly after the governed automatic path has already had enough access, …

    line 35
  9. Distinctive:

    If multiple blockers exist, preserve their ordering and do not hide the highest-severity one behind a degraded follow-up item.

    line 41
  10. Distinctive:

    Do not downgrade a live control-plane confirmation blocker into a generic degraded suggestion.

    line 42
  11. Distinctive:

    Treat unsupported platforms, unsupported Python versions, and missing editable-install availability as blockers.

    line 51
  12. Very specific:

    Treat only a self-contained prepared toolchain as ready for ordinary workspace asks.

    line 52
  13. Very specific:

    Treat missing uv, stale adapters, and an empty source corpus as degraded conditions rather than hard blockers.

    line 53

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Workspace Doctor

Use this skill when the task is to inspect the current workspace and explain what is missing or degraded.

Required Capabilities

  • local file access
  • shell or command execution
  • ability to summarize structured diagnostics

If the agent cannot inspect local files or run commands, stop and explain that the workspace cannot be diagnosed reliably.

Procedure

  1. Run docmason doctor --json.
  2. Explain blockers first, then degraded conditions, then optional follow-up.
  3. If the environment is not ready and .venv is absent, direct the next action to ./scripts/bootstrap-workspace.sh --yes.
    • treat the launcher as the governed cold-start preflight and bounded liveness gate for bootstrap Python selection, not as an unbounded trust of whatever python3 happens to be first on PATH
    • on native Codex, the ordinary path should not branch into manual cache probing or manual recovery before that governed launcher settles
  4. Otherwise, if the environment is not ready, direct the next action to docmason prepare --yes.
    • treat mixed and degraded toolchain states as repair-needed, not ordinary ask-time ready
  5. If doctor reports a control-plane blocker or pending confirmation, surface that blocker before lower-severity degraded conditions.
    • for pending native Codex host-access upgrade, direct the next action to switching the thread to Full access
    • for pending high-intrusion prepare, direct the next action to docmason prepare --yes
    • for pending material sync, direct the next action to docmason sync --yes
  6. If office-renderer is blocked:
    • on macOS with Homebrew already present, recommend brew install --cask libreoffice-still or docmason prepare --yes
    • on macOS without Homebrew, recommend docmason prepare --yes or the official installer from https://www.libreoffice.org/download/download/
    • on Linux, recommend the distro package manager or official packages, then re-run doctor
  7. If the normal launcher or prepare path still cannot complete honestly after the governed automatic path has already had enough access, point the deeper fallback to docs/setup/manual-workspace-recovery.md.
  8. Return the diagnosis to the main agent without mutating workspace state.

Escalation Rules

  • doctor is read-only. Do not silently switch into setup or repair work from inside this workflow.
  • If multiple blockers exist, preserve their ordering and do not hide the highest-severity one behind a degraded follow-up item.
  • Do not downgrade a live control-plane confirmation blocker into a generic degraded suggestion.

Completion Signal

  • The workflow is complete when the main agent has a clear blocker and degraded-condition summary plus the next obvious action.

Notes

  • doctor is read-only.
  • Treat unsupported platforms, unsupported Python versions, and missing editable-install availability as blockers.
  • Treat only a self-contained prepared toolchain as ready for ordinary workspace asks.
  • Treat missing uv, stale adapters, and an empty source corpus as degraded conditions rather than hard blockers.
  • On native Codex/macOS, treat missing LibreOffice as a blocker only when the workspace currently depends on Office rendering. Missing Homebrew alone is informational.
  • On native Codex/macOS, treat Default permissions versus Full access as an explicit first-class boundary; a per-command Yes popup is not equivalent to switching the thread to Full access.
  • Outside that native machine-baseline policy, treat missing LibreOffice as a blocker only when the current corpus contains PPTX, DOCX, or XLSX files.
  • For Claude Code or another compatibility host, keep any higher-access fallback wording brief and host-generic instead of copying the Codex-specific wording.
  • A stale or missing generated adapter matters only when the current flow depends on that adapter surface.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.