reviewing-release-runbook

Pass

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a workflow that ingests untrusted external data, which is a potential surface for indirect prompt injection attacks.
  • Ingestion points: In 'Step 1' of the workflow, the agent reads user-provided release runbooks and multiple upstream documents such as architecture and technical design files.
  • Boundary markers: The instructions do not mandate the use of delimiters or explicit 'ignore embedded instructions' warnings when processing the provided runbook content.
  • Capability inventory: The skill's primary capability is the generation of a textual 'approve' or 'revise' verdict and a list of findings; it does not explicitly perform code execution or external network calls.
  • Sanitization: There are no requirements for sanitizing or validating the input data format before the agent performs its analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 20, 2026, 08:02 PM
Security Audit — agent-trust-hub — reviewing-release-runbook