skills/mx-space/core/release-core/Gen Agent Trust Hub

release-core

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The prepare-server.sh script executes npx to download and run conventional-changelog-cli@5 from the npm registry.
  • [COMMAND_EXECUTION]: The skill runs shell scripts to perform Git operations (fetch, commit, tag, push) and Node/npm operations (pkg set, pnpm run, pnpm publish).
  • [DATA_EXFILTRATION]: The skill pushes code to the repository's origin and publishes packages to npm as part of the release cycle.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via Git commit messages ingested during the release notes generation process.
  • Ingestion points: preflight.sh reads commit history using git log.
  • Boundary markers: Absent in scripts; logic depends on agent summarization instructions in SKILL.md.
  • Capability inventory: The agent has permission to push to the remote repository and publish to the npm registry.
  • Sanitization: No sanitization is applied to the retrieved commit content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 08:44 PM
Security Audit — agent-trust-hub — release-core