The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to fetch and follow refactoring principles from an external website (loggingsucks.com) during the execution of the logging refactoring action. This external dependency introduces a surface where attacker-controlled content on the website could manipulate the agent's behavior during code modification.
Ingestion points: The references/logging.md file requires the agent to read loggingsucks.com in full to guide the implementation of wide event logging.
Boundary markers: The skill does not implement any boundary markers or instructions to ignore potentially malicious directions embedded in the external content.
Capability inventory: The skill has significant capabilities, including reading local source code, writing file modifications, and creating GitHub pull requests via the gh CLI.
Sanitization: There is no evidence of sanitization or validation of the content retrieved from the external website before it is used to influence the agent's refactoring logic.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to retrieve instructional data from a non-whitelisted third-party domain (loggingsucks.com) as a core part of its refactoring workflow.

performances