init

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted data from the repository's files to generate a persistent instruction file (AGENTS.md).
  • Ingestion points: The skill reads files including README.md, .cursor/rules/, .cursorrules, .github/copilot-instructions.md, and CLAUDE.md.
  • Boundary markers: No explicit boundary markers or warnings to ignore embedded instructions within the source files are provided.
  • Capability inventory: The skill possesses file read and file write capabilities (creating or modifying AGENTS.md).
  • Sanitization: No sanitization or validation of the ingested content is mentioned before it is processed and written into the new instruction file.
  • Risk: If a repository contains malicious instructions within its README or configuration files, those instructions could be incorporated into AGENTS.md, affecting the behavior of future agent sessions operating in that repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 04:37 PM
Security Audit — agent-trust-hub — init