Menu
Interactive guide for the SDD-AI methodology — explains which command to use based on what you need to do
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Guided complete flow to investigate and fix bugs — delegates investigation and fix to the refacil-debugger sub-agent in two passes separated by user confirmation
Internal reference — SDD-AI prerequisites shared by all other refacil skills (do not invoke manually)
Validate that the implementation meets the specs — builds a briefing with testCommand and CA/CR criteria, delegates to the refacil-validator sub-agent for the report, and handles corrections with user approval
Implement the tasks of a proposed change — verifies artifacts and working branch, builds a structured briefing, and delegates to the refacil-implementer sub-agent to execute the implementation in isolated context
Archive a completed change — move artifacts to archive and sync specs
Run the SDD implementation cycle autonomously after /refacil:propose was approved by the human — chains apply → test → verify → review → archive → up-code in a single invocation, and notifies the user via WhatsApp through Kapso when finished (success or failure). Use when the user says "autopilot", "ejecuta el resto del flujo", "termina solo", "modo autónomo", or indicates they will step away from the computer.
| name | refacil:guide |
| description | Interactive guide for the SDD-AI methodology — explains which command to use based on what you need to do |
| user-invocable | true |
You are a brief guide: you choose the next command; the detail of each flow is in the refacil-sdd-ai README (npm / CLI repo, section "Workflows") — do not repeat it word for word here.
propose → review artifacts → apply → test → verify → review → archive → up-code)./refacil:X?"/refacil:propose → (optional) /refacil:read-spec → then choose implementation:
/refacil:apply → /refacil:test (suite + coverage) → /refacil:verify (CA/CR; no full re-test by default) → /refacil:review (checklist; no suite) → /refacil:archive → /refacil:up-code/refacil:autopilot — chains apply → test → verify → review → archive in one invocation; up-code (push + PR) is optional and configured in pre-flight (the user chooses whether to end at archive or continue with up-code); optional WhatsApp via refacil-sdd-ai kapso setup/refacil:bug → /refacil:review → /refacil:archive → /refacil:up-code/refacil:explore/refacil:test/refacil:verify/refacil:review/refacil:up-coderefacil-sdd-ai init (global + per repo) and /refacil:setup/refacil:update/refacil:autopilot — same chain as option 1 path B; use when you want autonomous execution ("autopilot", "modo autónomo", "termina solo el flujo"). During pre-flight you define whether up-code (push + PR) is included — the cycle adapts and can end at archive or continue with up-code./refacil:read-spec — on-device browser playback; post-propose option B in propose, or on-demand for active changes / archived specs (refacil-sdd-ai read-spec --change <name>)/refacil:stats — task completion, review gate, test commands from memory.yaml (refacil-sdd-ai sdd stats <changeName>)/refacil:status — which phase of the SDD-AI cycle a change is in and the exact command to resume it (refacil-sdd-ai sdd status <changeName>)Note:
up-codeverifies.review-passedbefore push; seeMETHODOLOGY-CONTRACT.md §5-6for details.
Skill parity: 22 user-invocable skills (
user-invocable: true, excluding internalprereqs) must appear inSKILLS[], this menu, and the README "Available IDE Skills" table — includingread-spec,stats, andstatus.
stats vs status:
/refacil:statsshows telemetry (token savings, compact rewrites, CodeGraph calls, review history)./refacil:statusshows navigation (current cycle phase, next skill to run, state history). Use status to navigate the flow; use stats for observability data.
/refacil:propose is approved/refacil:read-spec (optional): opens the change folder in the browser and reads proposal, design, tasks, and specs aloud in order — useful before choosing implementation./refacil:apply (path A): step-by-step; each phase pauses for confirmation./refacil:autopilot (path B): fully autonomous; path B is independent — it runs review, archive, and optionally up-code internally. During pre-flight the user configures whether up-code is included (cycle ends at archive or at a PR). Does not merge into path A.For when the dev has multiple IDE sessions open (one per repo) and needs agents to consult each other without the dev being the manual transcriber:
| Command | When to use |
|---|---|
/refacil:join <room> | First step: joins this session to a room (creates the presentation block in AGENTS.md if missing). |
/refacil:say "..." | Announcement to the entire room. |
/refacil:ask @<repo> "..." [--wait N] | Query directed to another agent. With --wait N blocks until response or N sec. |
/refacil:reply "..." | Responds to the last question directed to this session. |
/refacil:attend | Puts this session in listening mode: when a directed question arrives, you answer and go back to listening. |
/refacil:inbox | View new messages since the last read. |
Typical pattern: before a task that may need context from other repos, the dev goes to the other windows and says "attend the bus". Then in their working repo, /refacil:ask @<other-repo> "..." --wait 180 brings the automatic response without jumping between windows.
Room agreements: if changes that touch this repo are agreed in the bus, the agent channels them with /refacil:propose (SDD-AI methodology), not loose patches unless the user explicitly orders otherwise. Whoever implements closes via bus with whoever requested the change (reply to the same ask thread, or ask/say according to refacil-prereqs/BUS-CROSS-REPO.md).
ask as request: if you request work in another repo, clear scope and criteria; they apply /refacil:propose there without repeating the guide (already in methodology via join). See /refacil:ask Step 1.5.
For monitoring: refacil-sdd-ai bus view (web UI) or refacil-sdd-ai bus watch <session> (terminal). No token consumption.
Full detail in the refacil-sdd-ai README (section refacil-bus).
/refacil:* commands in the chat or composer panel./refacil:read-spec or refacil-sdd-ai read-spec --change <changeName> — local TTS only, no remote APIs./refacil:autopilot [changeName] after propose is approved — in pre-flight you choose whether up-code (push + PR) is included. Configure Kapso once with refacil-sdd-ai kapso setup for WhatsApp notification on finish.Flow continuity: if the user confirms affirmatively ("yes", "ok", "go", "continue", etc.) to the continuity question in step 4 of Instructions, immediately execute the resolved /refacil:<skill> command. Deterministic resolution by Menu option:
/refacil:propose/refacil:bug/refacil:explore/refacil:test/refacil:verify/refacil:review/refacil:up-code/refacil:setupskill: "refacil:update"skill: "refacil:join" (or another from the group refacil:say/refacil:ask/refacil:reply/refacil:attend/refacil:inbox depending on the expressed intent)./refacil:autopilot/refacil:read-spec/refacil:stats/refacil:status/refacil:autopilot; if they want to hear specs first → /refacil:read-spec; if they want step-by-step → /refacil:apply.Do not describe it in text or wait for the user to type the command. (See METHODOLOGY-CONTRACT.md §5.)