production-deploy
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides a command to audit recent deployments by fetching Pull Request titles and bodies using the GitHub CLI (
gh pr list). This pattern ingests untrusted external data into the agent's context, which could theoretically contain adversarial instructions aimed at influencing the agent's behavior during the audit process. - Ingestion points:
SKILL.md(lines 92-94) fetches metadata directly from the GitHub API. - Boundary markers: The command does not use specific delimiters to wrap or isolate the untrusted PR text from the agent's core instructions.
- Capability inventory: The skill environment includes access to shell execution (
gh,git,kubectl,gcloud), package managers (pip), and the local filesystem. - Sanitization: Data is filtered through
jqto select specific fields, but there is no explicit sanitization of the text content itself.
Audit Metadata