flag-and-release-change

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the git CLI to perform repository operations including git clone, git fetch, git diff, and git push. These commands are used to manage pull request branches and land feature flag wiring in the codebase, which is the primary function of the skill.
  • [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests and analyzes pull request diffs, which are external and untrusted data sources.
  • Ingestion points: Untrusted data enters the agent context through the git diff origin/HEAD...<head_sha> command and the reading of changed source files during the 'Plan Phase' (SKILL.md).
  • Boundary markers: The instructions do not specify any delimiters or safety warnings to ignore instructions embedded within the processed code diffs.
  • Capability inventory: Across its operations, the skill can write to the local file system, commit and push to remote Git repositories, and perform LaunchDarkly API operations (via composed skills like launchdarkly-flag-create).
  • Sanitization: No sanitization, escaping, or validation of the PR content is described before the agent processes it for design and implementation decisions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 12:56 AM
Security Audit — agent-trust-hub — flag-and-release-change