skills/birdsigh/skills/agent-docs/Gen Agent Trust Hub

agent-docs

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to synthesize instructions for future agent sessions (stored in AGENTS.md) by surveying the current repository's code, git history, and documentation. This creates a surface for indirect prompt injection where malicious instructions hidden in project artifacts could be promoted into the agent's persistent instructional context.
  • Ingestion points: Project root files, package manifests, git logs, and specification/PRD documents (SKILL.md).
  • Boundary markers: The instructions do not define delimiters or provide specific warnings to the agent to ignore instructions embedded within the data being surveyed.
  • Capability inventory: The skill performs file system write operations to create documentation files (README.md, AGENTS.md, CLAUDE.md) that influence agent behavior in subsequent sessions.
  • Sanitization: There is no evidence of sanitization, filtering, or validation of the content extracted from external documents before it is written to the new instruction files.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 03:13 PM
Security Audit — agent-trust-hub — agent-docs