icp-builder
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process potentially untrusted external data, such as sales call transcripts and win/loss notes, which creates a surface for indirect prompt injection.
- Ingestion points: Input data specified in the 'What you need' section of
SKILL.md, including call recordings, transcripts, and customer descriptions. - Boundary markers: The instructions lack explicit boundary markers or delimiters for user-provided transcripts, which could theoretically allow embedded malicious instructions in those transcripts to influence the agent.
- Capability inventory: The skill produces a markdown file; no active tools (network, subprocess, or execution) are defined in the analyzed file.
- Sanitization: No sanitization or validation logic is present to filter malicious patterns from the input data.
Audit Metadata