release-announcement

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses several shell commands to extract repository information and build/preview the release site.
  • Evidence: Execution of git diff, git log, bun run release:announcement, ax ingest, ax sessions, ax recall, and bun run build as described in SKILL.md.
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it processes untrusted data from commit history and session logs.
  • Ingestion points: External data enters the agent's context through git log, git diff, and ax recall outputs (SKILL.md).
  • Boundary markers: The instructions do not specify any delimiters or warnings to the model to ignore embedded instructions within the ingested data.
  • Capability inventory: The agent can execute shell commands via bun, git, and ax, and write to the local file system (SKILL.md).
  • Sanitization: There is no evidence of sanitization or validation of the commit data before it is processed by the language model.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 05:08 AM
Security Audit — agent-trust-hub — release-announcement