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
- patricio0312rev/workspacestheir own version, first seen Jan 25, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- patricio0312rev/workspaces★ 9same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository17 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .claude/CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- adapter-syncSkill
- askSkill
- grounded-answerSkill
- grounded-compositionSkill
- knowledge-base-syncSkill
- knowledge-constructionSkill
- provenance-traceSkill
- retrieval-workflowSkill
- runtime-log-reviewSkill
- validation-repairSkill
- workspace-bootstrapSkill
- workspace-statusSkill
- operator-evalSkill
- public-sample-workspaceSkill
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
- Distinctive: line 8
Use this skill when the task is to inspect the current workspace and explain what is missing or degraded.
- Distinctive: line 16
If the agent cannot inspect local files or run commands, stop and explain that the workspace cannot be diagnosed reliably.
- Very specific: line 22
If the environment is not ready and .venv is absent, direct the next action to ./scripts/bootstrap-workspace.sh --yes.
- Very specific: line 23
treat the launcher as the governed cold-start preflight and bounded liveness gate for bootstrap Python selection, not as an unbounded trust …
- Very specific: line 26
treat mixed and degraded toolchain states as repair-needed, not ordinary ask-time ready
- Very specific: line 27
If doctor reports a control-plane blocker or pending confirmation, surface that blocker before lower-severity degraded conditions.
- Very specific: line 31
If office-renderer is blocked:
- Very specific: line 35
If the normal launcher or prepare path still cannot complete honestly after the governed automatic path has already had enough access, …
- Distinctive: line 41
If multiple blockers exist, preserve their ordering and do not hide the highest-severity one behind a degraded follow-up item.
- Distinctive: line 42
Do not downgrade a live control-plane confirmation blocker into a generic degraded suggestion.
- Distinctive: line 51
Treat unsupported platforms, unsupported Python versions, and missing editable-install availability as blockers.
- Very specific: line 52
Treat only a self-contained prepared toolchain as ready for ordinary workspace asks.
- Very specific: line 53
Treat missing uv, stale adapters, and an empty source corpus as degraded conditions rather than hard blockers.
≈ “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.
Files
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
- Run
docmason doctor --json. - Explain blockers first, then degraded conditions, then optional follow-up.
- If the environment is not ready and
.venvis 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
python3happens to be first onPATH - on native Codex, the ordinary path should not branch into manual cache probing or manual recovery before that governed launcher settles
- treat the launcher as the governed cold-start preflight and bounded liveness gate for bootstrap Python selection, not as an unbounded trust of whatever
- Otherwise, if the environment is not ready, direct the next action to
docmason prepare --yes.- treat
mixedanddegradedtoolchain states as repair-needed, not ordinary ask-time ready
- treat
- If
doctorreports 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
- for pending native Codex host-access upgrade, direct the next action to switching the thread to
- If
office-rendereris blocked:- on macOS with Homebrew already present, recommend
brew install --cask libreoffice-stillordocmason prepare --yes - on macOS without Homebrew, recommend
docmason prepare --yesor the official installer fromhttps://www.libreoffice.org/download/download/ - on Linux, recommend the distro package manager or official packages, then re-run
doctor
- on macOS with Homebrew already present, recommend
- If the normal launcher or
preparepath still cannot complete honestly after the governed automatic path has already had enough access, point the deeper fallback todocs/setup/manual-workspace-recovery.md. - Return the diagnosis to the main agent without mutating workspace state.
Escalation Rules
doctoris 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
doctoris read-only.- Treat unsupported platforms, unsupported Python versions, and missing editable-install availability as blockers.
- Treat only a
self-containedprepared 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 permissionsversusFull accessas an explicit first-class boundary; a per-commandYespopup is not equivalent to switching the thread toFull 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.
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
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.

