The Agent Skills Directory

[SAFE]: No security concerns were identified during the analysis of the skill's instructions, references, or assets.
[COMMAND_EXECUTION]: The skill executes a bundled Python script (assets/speccheck.py) to perform deterministic analysis. The script uses standard libraries and performs safe file operations using pathlib and regex-based parsing without calling external shell commands or using dynamic execution features like eval.
[DATA_EXFILTRATION]: The skill does not possess network capabilities and is explicitly designed as a read-only tool. It processes local project files only.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests user-controlled data (Go code and YAML specs), the risk is minimal. The deterministic script extracts structural data (names and types) and the LLM verification is scoped specifically to comparing these elements against defined specifications, following a strict verification protocol provided in the skill's references.

openapi-doc