go-to-market-strategy
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes information from external files such as
/foundation/brain.mdand/context/meta-patterns.md, as well as optional connectors like CRM and market data tools. This creates a surface for indirect prompt injection if these sources are populated with adversarial content. - Ingestion points: Data is loaded from
/foundation/brain.md(Sections 2, 3, 6),/context/meta-patterns.md, and optional external connectors defined in the SKILL.md Connectors section. - Boundary markers: The skill lacks explicit delimiters or instructions for the model to ignore potentially malicious instructions embedded within the ingested context files.
- Capability inventory: The skill is capable of reading local project files (Steps 0, 2) and appending session summaries to
/context/skill-sessions.md(Step 5). It does not exhibit dangerous capabilities such as arbitrary shell execution or unauthorized network exfiltration. - Sanitization: There is no evidence of sanitization, filtering, or validation performed on the data retrieved from the brain or connectors before it is processed.
Audit Metadata