The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands, including git diff to identify changes and a complex shell pipeline involving rg and sed to extract and verify documentation links across the project files.
[REMOTE_CODE_EXECUTION]: The skill directs the agent to dynamically identify and execute build or generation commands defined in project configuration files like package.json, CLAUDE.md, or AGENTS.md. This pattern allows for the execution of arbitrary local scripts if these configuration files are malicious or compromised.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by ingesting and acting upon untrusted data from the repository.
Ingestion points: Data is read from git diff output and markdown file contents via rg (ripgrep).
Boundary markers: No specific delimiters or safety warnings are used to isolate the processed data from instructions.
Capability inventory: The skill possesses capabilities for local command execution and interaction with the file system based on repository content.
Sanitization: The skill lacks sanitization or validation of the content extracted from the files before it is processed by the agent.

documentation-sync