antigravity-maintainer-batch-release

Pass

Audited by Gen Agent Trust Hub on Sep 24, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from external sources, including pull requests, issues, discussions, and entire skill subtrees during maintainer sweeps.
  • Ingestion points: The skill (specifically in the 'Maintainer Sweep' section of SKILL.md) reads data from open PRs, issues, and the full contents of skill directories (skills/**).
  • Boundary markers: The skill lacks explicit LLM boundary markers in its instructions but uses cryptographic fingerprints and head SHA attestations to track the integrity of reviewed data.
  • Capability inventory: The skill has the capability to execute repository scripts via npm run, perform Git operations, and interact with GitHub/npm APIs for merging and publishing.
  • Sanitization: It includes rigorous validation logic, such as byte-for-byte verification of generated outputs and mandatory semantic reviews, to mitigate the risk of processing malicious content.
  • [DYNAMIC_EXECUTION]: The skill constructs and executes shell commands for repository management tasks, including merging batches and preparing releases, using parameters supplied at runtime.
  • Evidence includes instructions to build command lines such as npm run merge:batch -- --prs <PR_LIST> and logic to ensure shell arguments are quoted when preparing installers to handle paths with special characters.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 24, 2026, 01:40 AM
Security Audit — agent-trust-hub — antigravity-maintainer-batch-release