tailrocks-improve
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary function is to perform automated repository audits using localized, read-only tools.
- [SAFE]: Implements explicit 'runtime-trust' rules that treat repository content as untrusted data and strictly forbid following embedded instructions (mitigating prompt/indirect injection).
- [SAFE]: Execution of verification commands is strictly gated by user authorization and sandboxing requirements (no network, secret scrubbing, read-only enforcement, time limits).
- [SAFE]: Uses a specific, internally installed command (
tailrocks.improve-report/v1) for report generation, ensuring controlled output. - [SAFE]: YAML frontmatter includes
disable-model-invocation: true, which prevents the model from autonomously calling tools unless explicitly triggered through the defined skill logic.
Audit Metadata