free-tools
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data such as scraped web text, user emails, and third-party exports. This represents a known attack surface for indirect prompt injection.
- Ingestion points: Data enters the context via external evidence sources defined in SKILL.md and references/skill-contract.md.
- Boundary markers: The skill mandates labeling all evidence with specific uncertainty tags (e.g., 'measured', 'estimated') to prevent truth-value confusion.
- Capability inventory: No risky capabilities such as arbitrary command execution, network exfiltration, or persistence mechanisms are requested or defined within the skill files.
- Sanitization: The skill contains explicit, high-priority instructions to treat scraped material as untrusted and strictly forbids following any instructions embedded within that data.
- [DATA_EXFILTRATION]: While the skill references marketing connectors (e.g., HubSpot, Google Ads) in references/connectors.md, it includes best-practice instructions to keep API keys out of source control and defaults to read-only behavior for scripts.
- [SAFE]: The skill adheres to the principle of least privilege, requiring explicit user authorization for any persistent writes, account changes, or external side effects.
Audit Metadata