development-estimation
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted data from the codebase and user-defined targets to generate complexity assessments and estimates, which creates a potential injection surface.
- Ingestion points: The
references/estimate.mdfile defines a custom command that takes a[target]parameter and utilizes an 'Explore' subagent to perform codebase complexity analysis. - Boundary markers: The skill lacks explicit boundary markers or instructions for the agent to ignore potentially malicious content found within the files it analyzes.
- Capability inventory: The skill has access to powerful tools including the
Tasktool (for delegating to subagents),Bash, and file system tools (Read,Grep,Glob), which could be targeted by an injection attack. - Sanitization: There is no evidence of input sanitization or output validation for the data retrieved from the codebase before it is used to drive subagent logic.
Audit Metadata