How do I install the Health Check skill?

Run npx skillmds@latest add tomevault-io/health-check-2 in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Health Check safe to use?

This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: PASS, Skill Scanner: PASS. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Health Check?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Health Check free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Health Check?

tomevault-io (@tomevault-io) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Health Check by tomevault-io — Agent Skill

Name: Health Check
Author: tomevault-io

Health Check

Periodic Codex repo health check orchestrator. Runs CI status, AGENTS.md staleness audit, documentation accuracy audit, and Obsidian Vault sync if present, pausing for approval between steps and producing one consolidated report. Global and project-agnostic. Trigger when the user says "health check", "health-check", "repo health", "check repo health", "repo maintenance", "run a health check", "audit the repo", "check everything", "periodic maintenance", or "check if everything is up to date". SKIP when the user only wants one specific audit. Use when this capability is needed.

tomevault-io Updated Sep 17, 2026

File contents

Repo Health Check

Run the maintenance sequence a Codex-supported repo needs: CI health, Codex instruction docs, user-facing docs, and knowledge base sync.

Sequence:

diagnose-ci -> audit-agent-docs -> audit-docs -> obsidian-vault-sync if a vault exists.

Operating Principles

Orchestrate, do not reimplement delegated skills.
CI health comes first.
Pause after each step and let the user skip any step.
Accumulate one consolidated final report.
Do not apply fixes automatically; each delegated skill owns its own approval gate.

Procedure

0. Orient

Confirm the cwd is the intended repo root.
Detect vault: check for <repo>/.obsidian-vault/.
Show the planned sequence and let the user drop steps.

1. CI Health

If GitHub Actions exists and gh is available, check recent run status.
If failing, invoke diagnose-ci to fetch logs, classify, and propose fixes.
If gh is unavailable, note the gap and continue.

2. Codex Instruction Docs

Invoke audit-agent-docs to score AGENTS.md and AGENTS.override.md files for staleness and coverage gaps.
Carry its findings into the consolidated report.

3. Project Documentation

Invoke audit-docs for user-facing and developer-facing docs.
Carry its findings into the consolidated report.

4. Obsidian Vault

If a vault exists, invoke obsidian-vault-sync with full scope unless the user narrows it.
If no vault exists, record "Skipped - no vault detected".

5. Consolidate

Return a table like:

Area              | Status | Findings
CI                | Clean  | 0 failing jobs
AGENTS.md audit   | Issues | N stale, M missing, P orphaned
Docs audit        | Issues | N high, M medium, P low
Vault sync        | Clean  | 0 findings

Then list next actions that still need user approval. Offer to save the report to .codex/health-check-<date>.md only after approval.

Guardrails

Never fabricate findings; use only delegated skill output and observed local state.
Do not silently skip unavailable skills; say what is missing.
Do not auto-apply fixes from delegated steps.
Keep the final report concise and actionable.

Source: ada-ggf25/AI-Tools — distributed by TomeVault.

---name: health-check-2description: Periodic Codex repo health check orchestrator. Runs CI status, AGENTS.md staleness audit, documentation accuracy audit, and Obsidian Vault sync if present, pausing for approval between steps and producing one consolidated report. Global and project-agnostic. Trigger when the user says "health check", "health-check", "repo health", "check repo health", "repo maintenance", "run a health check", "audit the repo", "check everything", "periodic maintenance", or "check if everything is up to date". SKIP when the user only wants one specific audit. Use when this capability is needed.---# Repo Health CheckRun the maintenance sequence a Codex-supported repo needs: CI health, Codex instructiondocs, user-facing docs, and knowledge base sync.Sequence:`diagnose-ci` -> `audit-agent-docs` -> `audit-docs` -> `obsidian-vault-sync` if a vaultexists.## Operating Principles1. Orchestrate, do not reimplement delegated skills.2. CI health comes first.3. Pause after each step and let the user skip any step.4. Accumulate one consolidated final report.5. Do not apply fixes automatically; each delegated skill owns its own approval gate.## Procedure### 0. Orient- Confirm the cwd is the intended repo root.- Detect vault: check for `<repo>/.obsidian-vault/`.- Show the planned sequence and let the user drop steps.### 1. CI Health- If GitHub Actions exists and `gh` is available, check recent run status.- If failing, invoke `diagnose-ci` to fetch logs, classify, and propose fixes.- If `gh` is unavailable, note the gap and continue.### 2. Codex Instruction Docs- Invoke `audit-agent-docs` to score `AGENTS.md` and `AGENTS.override.md` files for staleness and coverage gaps.- Carry its findings into the consolidated report.### 3. Project Documentation- Invoke `audit-docs` for user-facing and developer-facing docs.- Carry its findings into the consolidated report.### 4. Obsidian Vault- If a vault exists, invoke `obsidian-vault-sync` with full scope unless the user narrows it.- If no vault exists, record "Skipped - no vault detected".### 5. ConsolidateReturn a table like:```textArea | Status | FindingsCI | Clean | 0 failing jobsAGENTS.md audit | Issues | N stale, M missing, P orphanedDocs audit | Issues | N high, M medium, P lowVault sync | Clean | 0 findings```Then list next actions that still need user approval. Offer to save the report to`.codex/health-check-<date>.md` only after approval.## Guardrails- Never fabricate findings; use only delegated skill output and observed local state.- Do not silently skip unavailable skills; say what is missing.- Do not auto-apply fixes from delegated steps.- Keep the final report concise and actionable.---> Source: [ada-ggf25/AI-Tools](https://github.com/ada-ggf25/AI-Tools) — distributed by [TomeVault](https://tomevault.io).

Frequently asked questions