skills/project-status/SKILL.md
--- name: project-status description: Read-only status report that fuses the repo's handoff doc, live GitHub issues (incl. PRD umbrellas + their comments), the project's LLM-wiki vault, and git into ONE grouped roadmap — In-progress / Bugs / Blocked / Ready features / PRD umbrellas / Future — plus a dependency-ordered build queue of the ready, unblocked slices. The "where were we" answer when picking up a handoff. Writes the report to .handoff/status-latest.md so /ro:ultracode can orchestrate ag
npx skillsauth add RonanCodes/ronan-skills skills/project-statusInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
The "where were we" answer. A pure read-only synthesis across every durable source into one grouped roadmap a human can scan and an orchestrator can consume. It never commits, pushes, files issues, or edits code — the only thing it writes is the report file (skip with --no-write).
Sibling map:
/ro:session-state — the mechanical live-state sweep. This skill calls it as step 1./ro:handoff resume — re-enters a session and offers to dispatch the next move. This skill is the read-only report, not the dispatcher./ro:ultracode — consumes this skill's .handoff/status-latest.md build queue to orchestrate the work. This skill is the input; ultracode is the engine.Run /ro:session-state for uncommitted changes, unpushed commits, open PRs, in-flight branch. This is the "in-progress" truth.
# handoff / orientation doc (first hit wins)
for f in docs/agents/HANDOFF.md .handoff/latest.md HANDOFF.md docs/HANDOFF.md; do [ -f "$f" ] && echo "HANDOFF: $f" && break; done
git log --oneline -15
# GitHub backlog + PRD umbrellas + coordination comments
gh issue list --state open --limit 50 --json number,title,labels,createdAt 2>/dev/null
# read each `prd`-labelled issue body in full (it defines scope + slices)
# read the coordination-hub issue's comments (the in-flight / claimed-by anchor) if one exists
Read every prd-labelled issue body (the umbrellas define scope + remaining slices) and the coordination-hub issue's recent comments (what's claimed / in flight). Map each open enhancement issue to its parent PRD via the ## Parent\n\n#N convention or the slice lists in the PRDs.
If .ronan-skills-llm-wiki.local.json maps this repo to a vault, skim that vault's concepts/ for design docs the PRDs reference (e.g. makely-*.md). If no config and no vault, note "wiki: not mapped" and move on — don't block.
Read each issue's blocker signals (body, labels, comments, the handoff's BLOCKERS section) and sort into:
prd issues, each with: what's shipped vs what slices remain.Every item in 🚫 Blocked has a human action behind it. Re-frame each as a thing the user can go get, in if you want X → I need Y, from Z shape, so they can grab it without reading the whole report. Keep it to one line each by default:
🙋 What I need from you
- To get listing photos → an official feed deal (Realworks / brainbay), or a "yes, build the headless-image pipeline" call. Source: a vendor/legal decision, only you can make it.
- To make Saar call Dutch numbers (#27/#32) → a paid Vapi number or an imported Telnyx +31 number, then I swap VAPI_PHONE_NUMBER_ID. Source: Vapi/Telnyx dashboard, ~5 min.
- To send WhatsApp alerts (#13/#9) → Twilio sandbox keys in .env.local. Source: Twilio console.
Order by how much it unblocks (a decision that frees a whole PRD comes first). Mark each [decision], [key/secret], [spend], or [legal review] so the user sees the kind of action at a glance.
Default output is one line per ask (terse, grab-and-go). When the user says "elaborate on the blockers" / passes --elaborate, expand each into: what's blocked, why, the exact steps to get the thing, what it costs, and what unblocks the moment it lands. If the expanded list is long (4+ asks or multi-step procurement), also render it as a local HTML file and open it (never a remote Artifact — the user switches work accounts; this is a hard project rule), so they can work through it away from the terminal. Write to .handoff/what-i-need-<ts>.html.
From the Ready features only, emit an ordered list using the night-shift 4-signal intuition (age + priority + size + unblocks-others) and explicit deps (a slice that another slice needs comes first; respect PRD slice ordering). This is the machine-consumable handoff to /ro:ultracode.
Print the grouped roadmap to the user (markdown tables, emoji group headers, issue numbers as #N). Unless --no-write, also write the same report to .handoff/status-latest.md with a ## Build queue section as a checklist of - [ ] #N — <one-line> in order, so a downstream orchestrator reads a stable file. Create .handoff/ if missing (it's already gitignored in factory repos; if not, leave it untracked).
Always include the 🙋 What I need from you list (step 4b) — it's the payoff of the report. Keep it terse. Then end with two pointers when applicable:
→ say "elaborate on the blockers" and I'll expand each ask (+ a local HTML you can work off) when there is more than one ask.→ /ro:ultracode to fan out agents over the build queue when the queue is non-empty.Honour /ro:write-copy: no em/en-dashes, no AI-tell vocabulary, plain direct sentences. Group headers use emoji; product/UI copy never does (separate rule — this is a report, not product). Keep each item to one line.
# <Project>, project status (<date>)
🔨 In progress / next up — the active line + uncommitted/unpushed state
🐛 Bugs / tech-debt — table: item | notes
🚫 Blocked — table: issue | item | exact blocker
✨ Ready features — table: issue | item
🗺️ PRD umbrellas — per PRD: shipped vs remaining slices
🔮 Future / strategic — bullets
🙋 What I need from you — if you want X → I need Y from Z [decision|key|spend|legal]
## Build queue (ordered) — 1. #N — one-line (the /ro:ultracode input)
testing
--- name: linear-pipeline description: The Fable orchestrator for a single dispatched Linear ticket. Holds almost no context itself; it receives `--issue <ID> --detached`, decides the stage sequence, and fans out a sub-agent per stage, passing forward only each stage's artifact (never re-derived, never inlined into its own context). Step zero, before any planning or stage routing, is a boundary triage against `canon/security-boundary.md` (#199): a match tags Ronan Connolly and stops the run, no
development
--- name: in-your-face description: Capture a chat-only answer into a durable artifact (markdown + HTML, PDF when cheap) and launch it automatically so the user cannot miss it. Use when user says "in your face", "don't let me lose this", "save that answer", "make that durable", or right after answering a substantive side question (a recipe, comparison, how-to, or generated prompt) that would otherwise die with the context. category: workflow argument-hint: [--no-open] [--vault <short>] [hint of
tools
One-shot headless OpenAI Codex CLI calls for background/admin AI tasks — summaries, classification, extraction, admin glue. The default engine for anything that runs AI constantly in the background (daemon-driven, per-event), because it bills the flat ChatGPT subscription instead of Claude usage or per-token API spend, and it keeps working while Claude is rate-limited. NEVER for coding — coding stays Claude. Use when a skill or daemon needs a cheap always-on AI call, when the user says "use codex", "ask codex", "codex as backup", or when building a background summarizer/classifier into a listener or loop. Reads auth from ~/.codex/auth.json (ChatGPT account, no API key).
research
Turn a warranty rejection, repair quote, or RMA email into a cited decision brief — legal read (NL/EU consumer law), is the part user-serviceable, live part and new-unit prices, repair-vs-DIY-vs-new economics, before-you-send-it checklist, deadlines. Use when the user pastes or screenshots a repair quote, warranty rejection, "not covered" email, onderzoekskosten fee, or asks "should I repair or replace this".