deep
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements a mechanism to dynamically load and follow instructions from the SKILL.md files of other tools in the suite.
- Ingestion points: Instructions are retrieved from a specific "deep line" within the configuration files of tools such as drill, core-interview, implement, debug, verify, rubric, and land.
- Boundary markers: The skill does not define specific delimiters, validation rules, or "ignore instructions" warnings for the content it loads from these external files.
- Capability inventory: This specific skill (deep) does not invoke network, file-write, or shell execution tools directly, but it acts as a meta-instruction that alters the behavior of other tools which may possess those capabilities.
- Sanitization: No sanitization or filtering logic is described for the instructions dynamically retrieved from the suite's other configuration files.
Audit Metadata