The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is designed to generate and execute Python scripts (e.g., check.py) in the user's local environment to verify claims. This process involves the inherent risk of running dynamically generated code derived from external input.\n- [PROMPT_INJECTION]: The skill presents an indirect prompt injection attack surface by processing untrusted data from the draft results and dataset files. 1. Ingestion points: Data is ingested from the user-provided and files. 2. Boundary markers: The instructions do not mandate specific delimiters or "ignore instructions" warnings for the ingested content. 3. Capability inventory: The agent can execute shell commands, write files to the sandbox, and modify the report file. 4. Sanitization: No explicit sanitization or validation of the ingested data is performed before it is used to inform claim extraction or code generation.\n- [EXTERNAL_DOWNLOADS]: The instructions allow for the installation of external Python packages such as pandas via the uv package manager if standard libraries are insufficient. This involves downloading code from a public registry, although the skill requires explicit user consent for this operation.

claim-verify