the-list-builder
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from web searches and user-provided pastes. 1. Ingestion points: WebSearch and WebFetch tools used in SKILL.md and user-pasted exports from CRM or enrichment tools. 2. Boundary markers: Present; the skill includes a dedicated references/agent-security.md and explicit 'Agent Security' instructions in SKILL.md. 3. Capability inventory: The skill utilizes platform-provided WebFetch and WebSearch tools; it does not perform subprocess calls or local file-writes. 4. Sanitization: Instructions mandate quoting and reporting instructions found in content as findings rather than obeying them.
- [DATA_EXPOSURE]: The skill includes security best practices by instructing the agent to identify and flag secrets in input data (such as API keys in CRM exports) without reproducing or persisting them.
- [EXTERNAL_DOWNLOADS]: The skill accesses public web information through platform tools while explicitly prohibiting unauthorized scraping of gated services like LinkedIn, requiring user-provided paste paths instead.
Audit Metadata