action-runner

1# Action Runner Skill
2 
3The library is great at *recommending* work. This skill executes it — the action layer of the
4[Professional Brain](../professional-brain/SKILL.md) (Phase 2). A skill says "open a ticket per
5checklist item"; this turns that into real GitHub/Linear/Slack actions, **safely**: previewed,
6risk-rated, approved, then recorded. The cardinal rule: **nothing acts silently.**
7 
8## What This Skill Produces
9 
101. A **dry-run actions plan** — every proposed action with its target, operation, and **risk**.
112. After approval, the **executed actions** (via the connected action MCP) — outbound/destructive
12   ones gated individually.
133. A **record back to the brain** of what was actually done, with provenance.
14 
15## Required Inputs
16 
17Ask for (if not already provided):
18- **The recommendations to act on** (a launch checklist, PRD requirements, postmortem follow-ups…).
19- **The connected action MCP** and **targets** — which GitHub repo / Linear project / Slack channel. Scope is limited to what the user names; never act outside it.
20- **Approval posture** — what may run with a single OK vs. what needs per-action confirmation.
21 
22## How it works
23 
24```
25recommend → build an actions plan (JSON) → preview + risk-gate → approve → execute → record
26```
27 
281. **Build the plan** — express each action as JSON: `{"target","op","args","why","risk?"}`.
292. **Preview + gate** — run the helper; it prints a dry-run, classifies risk (🟢 low / 🟡 medium /
30…