m3-health

Solid

Health check — package version, installed payload, chatlog DB row count, per-agent hook state.

AI & Automation 16 stars 3 forks Updated today Apache-2.0

Install

View on GitHub

Quality Score: 77/100

Stars 20%
41
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
83
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

# M3 Health ## When to Use Use this skill when the user runs a system health check or when you want to diagnose issues with the m3-memory configuration, CLI executables, background services, or databases. ## Instructions Step 1 — run the doctor command, trying the resolvers below in order. Stop at the first one that returns exit 0; do not run the remaining ones. ```bash # 1. Plain CLI, if mcp-memory is on PATH: mcp-memory doctor # 2. Module form, works whenever the m3_memory package is importable: python -m m3_memory.cli doctor # 3. Repo-local venv (developer case, run from the repo root): .venv/Scripts/python.exe -m m3_memory.cli doctor # Windows .venv/bin/python -m m3_memory.cli doctor # macOS/Linux ``` Step 2 — print the full doctor output verbatim (no paraphrasing). Step 3 — append exactly ONE short line of interpretation. Examples: - `all healthy.` - `chatlog DB never captured — run /m3:install.` - `Antigravity SessionEnd hook off — run mcp-memory chatlog init --apply-gemini.` Do not write a paragraph. One line. The user can read the doctor output themselves. Step 4 — if (and only if) doctor reported a repairable problem (a stale/dead agent config path, a duplicate bridge, a disabled plugin, or a "run doctor --fix" hint), OFFER to auto-repair. `doctor --fix` repoints dead config paths, de-duplicates MCP registrations, and re-syncs agent hooks — it is non-destructive to data. Run the same resolver that worked above, with `--fix`: ```bash mcp-memory d...

Details

Author
skynetcmd
Repository
skynetcmd/m3-memory
Created
5 months ago
Last Updated
today
Language
Python
License
Apache-2.0

Integrates with

Bundled in these plugins

Similar Skills

Semantically similar based on skill content — not just same category