Renewal Risk Monitor

The name/description (renewal risk detection, structured brief output) match the included templates and scripts. The included Python script provides structured summarization and also supports directory/csv/pattern/skill audits — a broader capability than the minimal summary described in spec.json, but still coherent for audit/review workflows.

SKILL.md instructs the agent to either (a) run the local script with python3 or (b) generate output using bundled templates. The instructions do not ask the agent to access unrelated system credentials or remote endpoints. However, the script will recursively read many file types under any directory provided as input (markdown, json, py, sh, csv, etc.), so if the agent is pointed at a broad filesystem path it can surface arbitrary local files (including sensitive configs) in its output. SKILL.md does state '仅使用本地输入内容' and '默认只读', but operators must ensure the provided input path is limited to intended data.

No install spec; the skill is instruction-only plus a local Python script. It only requires python3 and uses the standard library. No downloads or external installers are invoked.

No environment variables, credentials, or config paths are requested. The script operates on user-supplied input files/directories. This is proportionate for a local auditing/reporting skill.

always:false and normal autonomous invocation settings. The skill does not request permanent platform-level privileges and does not alter other skills' configuration. It may write an output file if the operator allows it (standard behavior).