project-memory
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes repository files to generate agent-specific instructions, creating a surface for indirect prompt injection.\n
- Ingestion points: The skill reads existing
AGENTS.md,docs/, manifests, deployment/config files, tests, and scripts during the 'Setup or Refresh Workflow' described inSKILL.md.\n - Boundary markers: The skill instructions require the agent to 'Show a semantic summary and diff/full file before writing any AGENTS.md change' to the user for explicit approval.\n
- Capability inventory: The skill has the capability to write and update markdown documentation and instruction files (
AGENTS.md,docs/*) within the repository.\n - Sanitization: Mitigation is handled through mandatory human-in-the-loop review and approval before instruction-level files are modified.\n- [SAFE]: The skill implements security best practices by instructing the agent to exclude secrets, tokens, and sensitive environment data from the generated documentation, as specified in the 'Writing Rules' section of
SKILL.md.
Audit Metadata