Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add michaeljauk/brain-starter --skill lint-braingit clone --depth 1 https://github.com/michaeljauk/brain-starterWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/michaeljauk/brain-starter/lint-brain)<a href="https://agentmods.dev/skills/michaeljauk/brain-starter/lint-brain"><img src="https://agentmods.dev/badge/skills/michaeljauk/brain-starter/lint-brain.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00056 | $0.05089 |
| Opus 5 | $0.00028 | $0.02544 |
| Sonnet 5 | $0.00011 | $0.01018 |
| Haiku 4.5 | $0.00006 | $0.00509 |
Grade A, and why
lint-brain scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 448 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Lint Brain
Vault health check skill. Scans the brain repo for structural issues - orphan notes, broken links, missing metadata, stale files, and poorly connected knowledge islands. Reports a scored summary and optionally fixes issues.
Trigger phrases
/lint-brain- "lint the brain"
- "vault health check"
- "brain hygiene"
- "check vault health"
Vault location
~/brain/
Directories to check
projects/- project knowledge (one subdir per project, includes decisions)research/- non-project research (ai-agents, infrastructure, compliance, business, content)knowledge/- curated long-lived reference (lenny, people, companies)decisions/- cross-cutting decision recordsdocs/- repo meta, templates, skill output
Directories to skip
meetings/- read-only, managed by granola-sync. Never modify.sources/- gitignored third-party materialarchive/- inactive projects.obsidian/- Obsidian configtmp/- ephemeral files
Step 0 — run the stats script FIRST (do not count by hand)
python3 ~/brain/docs/scripts/lint/vault-stats.py --json
Every countable metric comes from this script. Orphans, broken links, missing frontmatter, staleness, near-duplicates, duplicate source URLs, thin-but-linked pages, knowledge islands, hubs, and existing contradiction flags are all computed deterministically in ~0.3s.
Do not recount them yourself. On 2026-08-02 this skill's first persisted audit got five counts wrong — broken links 311 vs ~120, near-duplicates 8 vs 0, stale manifests 4 vs 0, contradictions 2 vs 0, staleness 51 vs 542. This skill runs as a Haiku fork; arithmetic over 600 files is the wrong job for it. The script counts, you interpret.
What still needs you (and is not in the script):
- Check 13 — semantic contradiction detection. Requires reading and judgement; use
qmd query. - Check 5 — cross-link suggestions.
- Deciding which findings actually matter, and writing the analysis.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago First seen · 448 lines · 56 tokens per session scan A 02479a8d51d9
lint-brain is a skill published in the GitHub repository michaeljauk/brain-starter (5 stars, last pushed 11d ago), licensed MIT. It adds 56 tokens to every session and 5,089 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
log
Use this when a PersonalOS work session, chat outcome, personal reflection, or provided source should be persisted into the modular Daily context and any already-owned canonical files. Do NOT use for single-call processing, single-source knowledge ingestion, Gmail/WhatsApp propagation, or Lexware booking.
context-compression
A skill for condensing the history of meetings and projects into a short memory file. It records decisions, changes, outdated agreements, open questions, and key meetings.
parking
A workflow for saving the current state of a task before switching to another one. It records the stopping point, blockers, project plan changes, and information needed to resume later.
vault-onboarding-guide
An interactive guide for learning a file-based knowledge store with an agent. It teaches how to organize incoming material, projects, plans, tasks, logs, parked work, and resumed work.
new-dialog-handoff
A workflow for creating a new project in a knowledge vault, including its plan, tasks, context, activity log, and README. A knowledge vault is a structured collection of project notes and documents.
analyse-call
Use this when a local call transcript or meeting note should be analyzed into traceable PersonalOS context. Preserves the source, separates evidence from interpretation, updates only proven owner truth, routes confirmed commitments through task-manager, and never sends messages or triggers external actions without…