skills/session-state/SKILL.md
--- name: session-state description: The shared mechanical sweep that detects what's live in this repo/session but not yet durable — uncommitted changes, unpushed commits, open PRs, and whether the in-flight branch is tracked as an issue. Pure detection, no actions: it prints a structured inventory block that a calling skill acts on. Used by /ro:close-session ("I'm done") and /ro:handoff ("I'm continuing elsewhere") so the two never drift. Not usually invoked directly; triggers on "session state
npx skillsauth add RonanCodes/ronan-skills skills/session-stateInstall 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 mechanical durability sweep, factored out of /ro:close-session so that /ro:handoff runs the exact same detection. This skill only detects and reports — it never commits, pushes, files issues, or asks questions. The caller decides what to do with each finding.
/ro:close-session and /ro:handoff.Do NOT use it to resolve anything — it has no write tools by design. Resolution lives in the calling skill.
Run all four, quietly, then emit the inventory block.
git status --porcelain
git diff --stat
Report: file count + the --stat summary. Empty → "clean".
git log @{u}..HEAD --oneline 2>/dev/null
Report: the commit list (or "none" / "no upstream").
gh pr list --author @me --state open --json number,title,url,statusCheckRollup,mergeable 2>/dev/null
Report: per PR — number, title, CI rollup, mergeability. Skip silently if no gh remote.
git branch --show-current
gh issue list --search "<branch-keywords>" --state all --limit 5 2>/dev/null
Report: current branch, and whether an open issue plausibly tracks it (best-effort keyword match). Flag "branch not obviously tracked" when nothing matches.
--vault <name> or a vaults/<name>/ touch)sed -n '1,40p' vaults/<name>/ROADMAP.md
grep "^## \[" vaults/<name>/log.md | tail -3
Report: whether today's session appears in the log tail and whether ROADMAP has stale In progress rows. Detection only — /ingest session is the thing that actually updates them.
Emit exactly this block so callers can parse it consistently:
SESSION-STATE
branch: <name> (tracked-as-issue: yes #N | no)
uncommitted: <N files> — <stat summary, or "clean">
unpushed: <N commits, or "none">
open-PRs: <N> — [#N <title> · CI:<status> · mergeable:<state>] …
vault-drift: <"log missing today" / "roadmap stale" / "n/a">
─────────────────────────────────────────────
verdict: CLEAN-TO-CLEAR | HAS-LOOSE-STATE
verdict is HAS-LOOSE-STATE if any of uncommitted / unpushed / untracked-branch are non-empty, else CLEAN-TO-CLEAR. The caller reads the verdict to decide whether there's anything to resolve.
Two skills needed the same "what's not durable yet" detection: /ro:close-session (resolve it all, end clean) and /ro:handoff (capture it to a file, continue later). Duplicating the git/gh sweep in both meant they drifted. This is the single source of truth for detection; the callers own the acting.
[[close-session]] — resolves each finding inline; ends with nothing left.[[handoff]] — records each finding into one handoff file + ensures issues exist, then you clear.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".