The Agent Skills Directory

Dynamic Script Generation and Execution: The skill instructions direct the agent to generate and execute a Python helper script (workspace/helpers/append_review.py) to manage updates to finding files. This approach is used as a performance optimization to handle large JSON modifications, though it involves runtime code generation and execution.
Indirect Prompt Injection Surface: The skill ingests finding data from files in the workspace/findings/ directory. Since this data is generated by other agents or processes, the skill implements a mitigation strategy by instructing the agent to focus strictly on objective source code validation and disregard the qualitative reasoning provided in the input.
Command Execution: The skill utilizes shell commands to run its internal helper scripts and perform file system checks. These operations are localized to the defined workspace and target source code directories.

mantis-review