icp-fit-and-buying-readiness
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists entirely of markdown documentation and frameworks. No executable code, scripts, or automated tools are present in any of the skill files.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions involve processing untrusted data sources. 1. Ingestion points: CRM records, public company filings, and call transcripts (SKILL.md). 2. Boundary markers: Present; the skill strictly mandates that every output require cited evidence and that facts be separated from hypotheses (SKILL.md). 3. Capability inventory: None; the skill does not possess file-writing or shell-execution tools. 4. Sanitization: Present; the framework uses a diagnostic reasoning chain to ensure the agent remains grounded in verifiable observations rather than potentially adversarial instructions within the input data.
Audit Metadata