fixer
Pass
Audited by Gen Agent Trust Hub on May 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to discover and run the project's "canonical test command" from local files like the README, Makefile, or package scripts. This involves executing arbitrary shell commands found within the potentially untrusted repository.
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it processes external data to perform actions.
- Ingestion points: The agent reads "critical findings" from code reviews and traverses the project repository using tools like
rg(ripgrep). - Boundary markers: There are no markers or delimiters defined to isolate the agent's instructions from the content of the findings or the source code it processes.
- Capability inventory: The agent has the capability to modify any file in the repository and execute shell commands via the test verification step.
- Sanitization: No sanitization, validation, or filtering is applied to the findings or the code before the agent acts upon them.
Audit Metadata