| name | lint |
| description | Produce a read-only health report for the OKF second-brain knowledge bundle. |
| disable-model-invocation | true |
| allowed-tools | Read, Glob, Grep, Write, Bash(git diff *), Bash(git log *), Bash(git rev-parse *), Bash(git status *), Bash(git ls-files *) |
Lint Second Brain
Delegate structural and epistemic analysis to knowledge-curator.
Resolve scope from $ARGUMENTS in this order:
- If
governance/automation-state.md's last_successful_commit is empty, use full (even when $ARGUMENTS says incremental).
- Else if
$ARGUMENTS includes full, use full.
- Else if
$ARGUMENTS includes incremental, use incremental.
- Otherwise use
full (standalone /second-brain:lint default).
Callers that want incremental must pass the incremental token (maintain uses /second-brain:lint incremental verify).
For incremental:
- Collect candidates from
git diff --name-only <last_successful_commit> and from untracked/other knowledge paths via git status --porcelain / git ls-files --others --exclude-standard.
- Keep only
knowledge/**/*.md paths (plus one-hop knowledge Markdown links you discover from that set).
- If the filtered list is empty, write a no-op timestamped report under
governance/reports/ stating no knowledge pages changed since the baseline, and do not call the curator.
- Otherwise pass the filtered list to the curator with scope
incremental.
When $ARGUMENTS includes verify or strict, and a curator report was produced, also ask knowledge-verifier in audit mode to sanity-check that report. Skip audit for no-op empty-list runs.
Write only a timestamped report under governance/reports/. Do not modify knowledge pages, ledgers, or governance policy files. For every finding include severity, evidence, affected files, safe-versus-review-required classification (per governance/policies.md), quality-rubric notes when relevant, recommended action, and curator-required fields (including schema profile id when schema-related). Suggest research-scout for explicit gaps that need external candidates.
