contract-review
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill workflow is designed to process and analyze raw contract text provided by external sources, which is a known vector for indirect prompt injection.
- Ingestion points: Step 1 of the workflow and the example prompt explicitly instruct the agent to ingest raw contract text or excerpts provided by the user.
- Boundary markers: The skill does not provide instructions to use delimiters (such as XML tags or triple quotes) or specific warnings to ignore instructions embedded within the contract text.
- Capability inventory: The skill itself contains no scripts, tools, or direct capabilities for file system access, network operations, or command execution. It operates entirely as natural language instructions for the LLM.
- Sanitization: There are no instructions for the agent to filter or sanitize the input text for malicious prompts before processing.
Audit Metadata