recipe-fix-argparse-rejection
Pass
Audited by Gen Agent Trust Hub on Jul 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using
python3 .plan/execute-script.pywith the--helpflag. This is used for diagnostic purposes to retrieve valid script subcommands and flags from the script's own argparse configuration. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because its logic depends on processing potentially untrusted external data, such as script error outputs and documentation from other skills.
- Ingestion points: Argparse error strings from
stderrand the## Canonical invocationssection of externalSKILL.mdfiles. - Boundary markers: None specified to delimit or sanitize the external error messages or documentation content.
- Capability inventory: The agent is authorized to perform subprocess execution (via the help commands) and file modification (to fix the rejected call sites in workflow documents).
- Sanitization: No specific sanitization or validation of the input strings is mentioned before they are used to reconstruct shell commands or update files.
Audit Metadata