update-agent-docs

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injection Surface: The skill defines a process for updating the agent's knowledge base by reading repository changes and writing to memory files in .github/memory/ and .github/instructions/. This creates a potential surface where external content (like code comments or documentation updates from a pull request) could influence the agent's future behavior if that content contains embedded instructions. This is a common characteristic of documentation assistants.
  • Ingestion points: Processes code changes, public API modifications, and diagnostic updates as described in SKILL.md.
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore' directives to separate data from instructions during the update process.
  • Capability inventory: The skill grants the agent the ability to write to and evolve documentation files within the .github/ directory across all scripts.
  • Sanitization: There are no explicit steps defined for sanitizing or validating the content before it is committed to the agent's memory files.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 10:37 AM
Security Audit — agent-trust-hub — update-agent-docs