cite-or-stop
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions involve the agent ingesting and citing information from external and potentially untrusted sources including codebase files, issue trackers, and runtime command outputs.
- Ingestion points: Data is pulled from repository file paths, ticket management items (Jira/AC), and shell command outputs (SKILL.md).
- Boundary markers: The skill defines citation tags like
[code],[runtime], and[ticket]to distinguish source content, and uses theInvented (not approved):prefix to isolate unverified claims. - Capability inventory: The instructions assume the agent has the ability to read files, view ticket metadata, and execute commands to observe runtime state.
- Sanitization: The skill does not define specific sanitization or filtering logic for the data ingested from these external sources before it is processed by the agent.
Audit Metadata