catch-me-up
Skill0 starsChanged 17 days agoMITView on GitHub
Reconstruct what happened on a project while someone was away — decisions made, direction changes, things now blocked on them, and what is safe to ignore. Use after leave, a sprint spent heads-down, joining a project mid-flight, or when asked "what did I miss?".
4 other projects use a version of this.
Adopted by
Most recently taken up by
- khgouldy/skillstheir own version, first seen Nov 1, 2026
- t1merickson/skillstheir own version, first seen Nov 1, 2026
- palmeheskd/claude-skillstheir own version, first seen Oct 25, 2026
- RisorseArtificiali/skillstheir own version, first seen Sep 6, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 4
- RisorseArtificiali/skills★ 13same name
- khgouldy/skills★ 1same name
- t1merickson/skills★ 1same name
- palmeheskd/claude-skills★ 0same name
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 repository4 other files it gives its agents.
- action-item-auditSkill
- commitment-trackerSkill
- decision-archaeologySkill
- decision-recordSkill
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
4
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 53
Prefer under-reporting to a false "you're blocking X".
- Distinctive: line 57
Keep it to one screen. If it runs longer, you have included activity instead of change.
- Distinctive: line 85
Never invent a decision: If the record is ambiguous, say "discussed, no decision recorded" — that is a real and useful state.
- Distinctive: line 89
If the window contains nothing significant, say so in one line.
“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Catch me up
Returning after time away, the problem is never a lack of information. It is 400 unread messages, 60 commits and 12 meetings, with no signal about which 5% changed anything.
This skill produces a brief that respects the reader's time: what changed, what needs them, what they can skip.
Inputs you need
Establish both before starting, and ask if either is missing:
- Window — from when to when. "Last two weeks", "since Oct 3".
- Scope — which project, repo, team or space. Without this you will produce noise.
Procedure
1. Decisions first, activity last
Open with what was decided, not what was done. Activity is cheap; decisions are what the reader must align to.
With a Wisprnote MCP server connected:
list_meetingsacross the window to see the shape of what happenedget_meeting_knowledgeon each to pull decisions without reading whole transcriptslist_action_itemsfiltered to the reader's name — this becomes the "needs you" section
Otherwise: merged PRs with substantive descriptions, closed issues labelled decision or
RFC, ADRs added under docs/, changelog entries.
2. Find direction changes
These matter more than anything else and are the easiest to miss. Look for:
- A decision that reverses an earlier one
- Work abandoned mid-flight — branches gone stale, issues closed as won't-fix
- Scope changes — things added to or cut from a release
- Ownership changes — someone left, someone took over
A reversal the reader doesn't know about will have them confidently doing the wrong thing within an hour of returning. Lead with these.
3. Find what is blocked on them
Anything where the reader is named as owner, reviewer, or approver, and the thing is still open. This is the only section that generates immediate action, so it must be accurate. Prefer under-reporting to a false "you're blocking X".
4. Write the brief
Keep it to one screen. If it runs longer, you have included activity instead of change.
# Catch-up: <scope>, <window>
## Changed direction (read this first)
- <what changed, and what it means for work in flight>
## Decisions
- <decision> — <date>, <who> · <where it was decided>
## Blocked on you
- <item> — <who is waiting, since when>
## Worth knowing
- <2-5 items that don't need action but change your model of the project>
## Safe to skip
- <categories of noise, summarised in one line — not enumerated>5. Offer depth, don't deliver it unasked
End by naming the two or three threads where a deeper read would pay off, and let the reader pick. Do not pre-emptively expand all of them.
Guardrails
- Never invent a decision. If the record is ambiguous, say "discussed, no decision recorded" — that is a real and useful state.
- Attribute carefully. Getting "who decided" wrong is worse than omitting it.
- "Safe to skip" is a summary, not a list. Enumerating noise recreates the problem.
- If the window contains nothing significant, say so in one line. A short honest brief is a good outcome, not a failed one.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
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.

