demo-guide
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the project's filesystem, including configuration files, source code, and specifications, which could potentially contain malicious instructions.
- Ingestion points: The skill reads
docs/project-config.json, source code files, and specification documents as identified in Step 0.2 and Step 2. - Boundary markers: The skill does not define specific prompt delimiters for external content but relies on
file:lineverification as a control to ensure evidence-based results. - Capability inventory: The skill is capable of writing documentation to the local filesystem (
docs/demo-guides), spawning sub-agents for complex tasks, and executing local support scripts. - Sanitization: A strict redaction policy is enforced, where credentials and tokens are replaced with
<redacted:…>placeholders before they enter the processing context (Step 2). - [COMMAND_EXECUTION]: The skill provides instructions for the agent to optionally execute a local Python script (
.claude/scripts/code_graph) to assist with code tracing and blast radius analysis. This script is part of the local environment provided by the vendor (Step 2).
Audit Metadata