nio-actionlisted

# Nio — Action Safety Evaluation Evaluate whether a proposed runtime action should be allowed, denied, or require confirmation. This is the focused `action` capability of the Nio framework. For detailed policies and detector rules, see [ACTION-POLICIES.md](ACTION-POLICIES.md). > **Passive invocation.** If the user asks whether some concrete command/request/file-op/secret-access is safe to run, you MUST evaluate it with the CLI below rather than guessing. (For scanning a body of code/files, use `nio-scan`; for live endpoint scores, use `nio-external-score`.) ## Resolving the Script Path The CLI for this skill lives in the **sibling `nio` skill**, not in this skill's own directory. Resolve it relative to where you loaded this SKILL.md: 1. This SKILL.md's parent directory is THIS skill's directory (e.g. `<plugins>/skills/nio-action/`). 2. The script is the sibling `nio` skill's `scripts/action-cli.js` — i.e. `../nio/scripts/action-cli.js` relative to this directory. Derive the absolute path; do **not** hard-code `~/.claude/...`. 3. Invoke with a **single** `node` command — no `cd`, no `;`/`&&`/`|`/`$(...)`/backticks. Example (substitute the real absolute path): ```bash node /absolute/path/to/skills/nio/scripts/action-cli.js decide --type exec_command --command "ls" ``` ## Supported Action Types - `network_request` — HTTP/HTTPS requests - `exec_command` — Shell command execution - `read_file` / `write_file` — File system operations - `secret_access` — Environment variabl