The Agent Skills Directory

[DYNAMIC_CONTEXT_INJECTION]: The skill uses the !command`` pattern in the Context section to execute find commands. These commands are used to identify relevant file types within the project. These are benign, read-only operations intended to provide the agent with situational awareness.\n- [COMMAND_EXECUTION]: The skill utilizes bash to run ast-grep and ripgrep. Safety is maintained by the allowed-tools configuration in the YAML frontmatter, which restricts bash usage specifically to the sg and rg commands, preventing the execution of arbitrary or malicious shell commands.\n- [INDIRECT_PROMPT_INJECTION]: The skill has an ingestion surface as it reads and analyzes codebase files which may contain untrusted content.\n
Ingestion points: Files are read using Read, Grep, and ast-grep (SKILL.md).\n
Boundary markers: Not explicitly defined for the content of the files being analyzed to separate code from potential embedded instructions.\n
Capability inventory: Limited to Bash(sg/rg), Read, Glob, Grep, TodoWrite, and Task (SKILL.md).\n
Sanitization: No specific sanitization or filtering of external content is mentioned. The risk is limited to influencing the analysis report generated by the agent.

code-antipatterns