ai-estimate
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill methodology directs the agent to ingest and analyze data from external, untrusted sources including project tickets, Slack/Teamwork threads, and pull request comments.
- Ingestion points: External communication channels such as tickets and messaging threads referenced in the procedure section of SKILL.md.
- Boundary markers: The instructions do not specify any delimiters or safety warnings to distinguish external content from agent instructions.
- Capability inventory: The grounded estimation process requires the agent to read local codebase files.
- Sanitization: No methods for sanitizing or escaping the external content are included in the skill's instructions.
Audit Metadata