revealui-pickup
Skill2 starsChanged 40 days agoMITView on GitHub
Consume the rolling CURRENT-HANDOFF and continue agent-doable work. Use when the user runs /pickup, asks to continue from checkpoint, pick up the last session, or resume fleet handoff work in a new session. Not /next (TRACKER free surfaces). Not /resume (vendor transcript). Not /recover (crash inventory).
We split this skill into the rules it gives an agent
12 rules. About 0 are common, 7 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Distinctive: line 31
Do not run this on SessionStart.
- Distinctive: line 52
If the file is missing: say so, run Step 6 (TRACKER print), stop.
- Distinctive: line 53
If Launch is missing: fail closed — name that the fragment predates ADR 2026-08-26-session-launch-record.
- Very specific: line 93
If any agent-doable remains after Launch: do it this session.
- Very specific: line 101
Never merge, force-push, add gate labels, or edit a stranded .jv checkout (docs/gap-381-yml-with-render or any chore/checkpoint-* main …
- Distinctive: line 126
Wait for the owner to pick.
- Very specific: line 130
Do not steal /next or print TRACKER before finishing Steps 1–4.
- Distinctive: line 131
Do not paste a next-agent prompt back at the owner.
- Very specific: line 132
Do not start /recover unless git is corrupt or the owner asked.
- Distinctive: line 133
Do not auto-run on SessionStart.
- Distinctive: line 134
Do not guess a product when Launch is missing.
- Very specific: line 135
Do not rewrite tracked CURRENT-HANDOFF.md on the shared primary .jv.
≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Pickup (/pickup) — consume CURRENT-HANDOFF
Clean continue after /checkpoint. Read the rolling handoff, re-verify live
git/gh, then do remaining agent work. Do not invent a second board.
/next stays TRACKER free surfaces. /recover stays crash diagnostics.
Grok /resume and grok --continue reload a transcript, not this file.
Do not run this on SessionStart. Opening a session is not consent to continue last night's PR.
Load helpers:
. "$REVEALFLEET_ROOT/revskills/scripts/lib/session-state.sh"Step 1 — Render + read
CURRENT_HANDOFF="$(cd "$JV_REPO" && node scripts/handoff-render.js)"Stdout is the path to read. On the shared primary checkout the renderer
writes gitignored docs/handoffs/.CURRENT-HANDOFF.local.md so .jv stays
clean. In a worktree it writes CURRENT-HANDOFF.md in that tree. Do not
commit the render. Then read $CURRENT_HANDOFF sections Launch,
Ordered next actions, Owner-gated, In-flight, and the newest
rolling fragment under docs/handoffs/rolling/.
If the file is missing: say so, run Step 6 (TRACKER print), stop. If Launch is missing: fail closed — name that the fragment predates ADR 2026-08-26-session-launch-record. Do not guess a product.
Step 2 — Re-verify (mandatory)
Fragments go stale the moment the owner merges. Do not trust prose.
Classification is handoff-status.js, the same script the "what now?"
status line uses. Do not re-implement gh pr view in the session.
node "$JV_REPO/scripts/handoff-status.js"done is merged or closed. owner-gated is an open pull request whose
remaining line is an owner merge, or a pull request named only under
Owner-gated. agent-doable is an open pull request whose ordered-next
line still says to clear, fix, push, or re-run CI.
Also confirm named worktrees/branches still exist:
cd "$REVEALFLEET_ROOT/revealui" && git worktree listStep 3 — Classify
| Class | Meaning | This session |
done | Live git/gh matches the tail (merged, deleted, already applied) | Note and skip |
owner-gated | Merge, vault, prod, promote, machine UAC, stranded checkout | List the one-line owner command. Do not do it |
agent-doable | Spec, PR, tests, docs, conflict fix, follow-up an agent can do | Continue it |
Print a short live board. Stale fragment lines stay in the file; live state wins.
Step 4 — Continue
First action is CURRENT-HANDOFF ## Launch row 1 (exact rfg/rfc
command, including --worktree=). That is Ordered next item 1. Then:
- If any
agent-doableremains after Launch: do it this session. Worktree fromorigin/testin the product named by Launch. Do not dirty-switch a shared checkout. - Else if only
owner-gatedremains: list the script'sgh pr mergelines, then run Step 6. Do not merge. When the owner says "merged", re-runhandoff-status.jsand continue Step 4. - Else (checkpoint exhausted): Step 6.
Never merge, force-push, add gate labels, or edit a stranded .jv checkout
(docs/gap-381-yml-with-render or any chore/checkpoint-* main HEAD).
Step 5 — Report
=== PICKUP ===
Handoff: $CURRENT_HANDOFF
Live: <one line: what is still true after gh>
Doing: <agent-doable item 1 | owner-gated wait | tracker fallthrough>
Owner-gated: <commands or none>Then execute the Doing line. On owner-gated, print the merge lines and Step 6. Do not merge.
Step 6 — Board under the handoff
Same entry as /next. sync refreshes the board and the goals overlay.
tracker.js next does not.
node "$JV_REPO/scripts/tracker.js" syncPrint free surfaces under the handoff board. Do not auto-claim a gap. Wait for the owner to pick.
Do not
- Do not steal
/nextor print TRACKER before finishing Steps 1–4. - Do not paste a next-agent prompt back at the owner; this skill is the consume path.
- Do not start
/recoverunless git is corrupt or the owner asked. - Do not auto-run on SessionStart.
- Do not guess a product when Launch is missing.
- Do not rewrite tracked
CURRENT-HANDOFF.mdon the shared primary.jv. - Residue after consume is
/cleanup(report) or/cleanup --fix(gated). Stale MASTER_HANDOFF is/rollup. Do not invent a second sweep.
Other versions1 project uses a version of this.
Word for word · 1
- RevealUIStudio/revskills★ 2100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository29 other files it gives its agents.
- skills/revealui-design/CLAUDE.mdCLAUDE.md
- drizzle-dbSkill
- electric-syncSkill
- exhaustive-auditSkill
- grok-bot-checkpointSkill
- knowledge-graphSkill
- mcp-serverSkill
- multi-agent-memorySkill
- next-best-practicesSkill
- redundancy-scanSkill
- revealui-checkpointSkill
- revealui-cleanupSkill
- coordinateSkill
- revealui-designSkill
- revealui-design-statusSkill
- revealui-doctorSkill
- revealui-opsSkill
- revealui-recoverSkill
- revealui-rollupSkill
- revealui-skills-testSkill
- revealui-snapshotSkill
- revealui-sync-ltsSkill
- revealui-sync-rulesSkill
- revealui-trackerSkill
- revvault-resolveSkill
- security-hardeningSkill
- tailwind-v4Skill
- vitest-testingSkill
- yjs-collaborationSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

