code-reading
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill framework is designed to ingest and analyze external source code, which introduces a surface for indirect prompt injection from untrusted data.
- Ingestion points: The skill reads file contents and directory structures as part of the Acquisition layer defined in SKILL.md and layers/acquisition.md.
- Boundary markers: The provided instructions and templates do not include delimiters or specific instructions for the agent to ignore embedded prompts when processing code.
- Capability inventory: The skill uses file system listing and reading operations, and it has the capability to delegate tasks to sub-agents for exploration.
- Sanitization: No sanitization, escaping, or validation of the ingested code content is performed.
Audit Metadata