kentcdoddskentcdodds/kcd-skills/skills/visual-recap/SKILL.md

visual-recap

Skill64 starsChanged 24 days agoMITView on GitHub

When a change is non-trivial (a plan, a PR create or update, or a request for a visual recap), put one marker-delimited block in the PR description. GitHub renders it. It does not replace reading the diff. Recap mode reads the diff, not memory. Claims about what changed must be checkable against git diff <base>...HEAD (plus --stat). Session context may explain intent only. Run the classifier. Roll up to the highest of adds, then extends, then composes. You still choose…

8 other projects use a version of this, 2 of them edited.

Adopted by

Most recently taken up by

chrislacey89/skillstheir own version, first seen Nov 4, 2026
BuilderIO/skillstheir own version, first seen Oct 28, 2026
ericmasiello/skillsedited, first seen Oct 20, 2026
malinskibeniamin/skillstheir own version, first seen Oct 17, 2026
willlee1995/skillstheir own version, first seen Oct 13, 2026
tslateman/skillstheir own version, first seen Oct 11, 2026

Other versions8 projects use a version of this.
Word for word · 0
None found.
Edited · 2
kentcdodds/kody★ 92681% same
ericmasiello/skills★ 011% same
Wrote their own version · 6
BuilderIO/agent-native★ 7.6ksame name
BuilderIO/skills★ 4.7ksame name
chrislacey89/skills★ 4same name
malinskibeniamin/skills★ 4same name
tslateman/skills★ 1same name
willlee1995/skills★ 1same 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 repository6 other files it gives its agents.
- autopilotSkill
- friction-logSkill
- prefer-local-kody-executeSkill
- review-and-recommendSkill
- ship-prSkill
- use-kodySkill

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 2 are distinctive, and 3 are very specific to this project.

0
Common
Most of the 8 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Distinctive:
When a change is non-trivial (a plan, a PR create or update, or a request for a visual recap), put one marker-delimited block in the PR …
line 15
Very specific:
Run the classifier. Roll up to the highest of adds, then extends, then composes.
line 25
Very specific:
Update primitives.yaml when this PR adds, removes, or materially reshapes a primitive (new id, renamed meaning, or ownership roots that …
line 35
Very specific:
Keep each entry's adrs in sync: list the current decision records that shape the primitive (adrs: [] if none), and when this PR adds or …
line 40
Distinctive:
Do not cat the session cookie into curl or Python.
line 77

“Common” is measured: kept by at least 6 in 10 of the 8 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.

Keep what fits and drop the rest. Your version, ready for your agents.

Files

SKILL.md

System recap

When

Recap mode reads the diff, not memory. Claims about what changed must be checkable against git diff <base>...HEAD (plus --stat). Session context may explain intent only.

Run the classifier. Roll up to the highest of adds, then extends, then composes. You still choose composes vs extends from the diff. adds is a new map entry.

Rollup	Risk	Also
`composes`	low	Wiring and call sites only.
`extends`	medium	Verify locally with the project's control CLI as the seeded user; a PR preview only when local cannot
`adds`	high	Update `primitives.yaml`, run `check-primitives-map.mjs`, and the same local verify (preview when needed)

Update primitives.yaml when this PR adds, removes, or materially reshapes a primitive (new id, renamed meaning, or ownership roots that must change). Do not edit summary for ordinary feature work. Call out a touched invariant from that map at every risk.

Keep each entry's adrs in sync: list the current decision records that shape the primitive (adrs: [] if none), and when this PR adds or supersedes a record, add it to (or swap it into) the matching entries. The classifier prints each touched primitive's adrs; read them before you change that primitive. The check script fails on missing or superseded paths.

Plan mode uses **Mode:** plan before a PR exists. Recap mode replaces that block from git diff <base>...HEAD. In plan mode, say explicitly when the plan requires no change to any primitive; that is the lowest-risk outcome.

The block shape is one example: references/block-format.md.

Command

node .agents/skills/visual-recap/scripts/classify-primitives.mjs --base <base> --head HEAD --json
node .agents/skills/visual-recap/scripts/upsert-recap-block.mjs <pr-number> <block-file>
node .agents/skills/visual-recap/scripts/check-primitives-map.mjs

The upsert script checks mermaid, then replaces the marker block or appends it. It does not edit text outside the markers. Re-run both commands after a significant push.

Failure

The mermaid check exits non-zero. Fix the diagram. Keep ; out of sequence notes and messages.
The block file must start with  and end with .
gh pr edit fails with Resource not accessible by integration (updatePullRequest). The script still prints the merged body. Apply that body with Cursor ManagePullRequest. Do not have Kody edit the PR.
A health check or login smoke is not medium or high verification. Prefer local control CLI proof; use a preview only when local cannot exercise the change. Do not cat the session cookie into curl or Python.

Download now (ZIP)

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

What did you use it for? e.g. Reviewing pull requests in a Rails app

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.

Check your setup, free