ship-learn-next
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted content such as YouTube transcripts, articles, and notes using the Read tool to extract implementation steps. This content is then used to generate a plan that is written to the filesystem using the Write tool. This creates a surface for indirect prompt injection where malicious instructions embedded in the source material could influence the agent's behavior. Evidence Chain: 1. Ingestion points: Reads user-provided files via the Read tool as described in Step 1 (SKILL.md). 2. Boundary markers: Absent; the instructions do not include delimiters or warnings to ignore instructions embedded in the source material. 3. Capability inventory: Uses the Read and Write tools to access and modify the local filesystem. 4. Sanitization: Absent; the skill lacks validation or filtering of the external content before processing.
Audit Metadata