he-runbook

Pass

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses local shell commands and scripts, such as find, rg (ripgrep), scripts/ci/he-runbooks-lint.sh, and scripts/runbooks/select-runbooks.sh. These are standard repository maintenance tools used here to inventory and validate documentation files.
  • [PROMPT_INJECTION]: The skill processes markdown files from the docs/runbooks/ directory, creating a surface for indirect prompt injection. This is addressed through a structured security framework:
  • Ingestion points: Files are read from docs/runbooks/*.md during the inventory and edit phases.
  • Boundary markers: The skill enforces specific frontmatter requirements and a mandatory "additive-only" declaration in the runbook body.
  • Capability inventory: The skill has the ability to read and write files within its designated directory and execute local validation scripts.
  • Sanitization: All content is audited against references/audit-rules.md, which contains strict prohibitions against gate waiving, evidence weakening, and unauthorized privilege escalation (e.g., force-pushing without consent).
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 20, 2026, 09:12 PM
Security Audit — agent-trust-hub — he-runbook