i2r-elicitation-mode
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [NO_CODE]: The skill consists entirely of markdown files and YAML metadata. It does not include any executable scripts (.py, .js, .sh), binaries, or automation commands. Behavior is governed strictly through natural language instructions.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing untrusted data from founder ideas and interview answers. Analysis of this surface:
- Ingestion points: Untrusted content enters via raw ideas (e.g.,
raw/idea.md) and clarification responses. - Boundary markers: The skill acts as a boundary enforcement mechanism by requiring explicit categorization of all input into
stated,assumed, ordecisionsbuckets. - Capability inventory: The skill possesses no dangerous capabilities, network tools, or system-level access; its purpose is restricted to generating structured analysis files.
- Sanitization: It implements a rigorous 'Mom Test Filter' which requires verifiable
source_refpointers for any 'stated' facts, effectively preventing unverified instructions embedded in user data from being treated as grounded requirements. - [SAFE]: No malicious patterns such as obfuscation, credential harvesting, or remote code execution were detected during the analysis.
Audit Metadata