The Agent Skills Directory

[SAFE]: The skill's primary function is managing documentation (specifically CODING_STANDARDS.md). It includes clear instructions to avoid modifying executable configurations, agent instructions, or application code without separate authorization.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by reading project files like CODING_STANDARDS.md and various configuration files. However, it incorporates a mandatory human-in-the-loop checkpoint ('Obtain acceptance' step) that requires explicit human approval of proposed deltas before any file writes occur, effectively mitigating the risk of automated obedience to embedded instructions.
[COMMAND_EXECUTION]: There are no shell commands, subprocess calls, or scripts included in the skill. It operates through natural language instructions for the agent.

coding-standards