summarize-change
Skill1 starsChanged 16 days agoMITView on GitHub
Summarize the changes of a Jujutsu revision, as high level prose.
In this repository15 other files it gives its agents.
- audit-skill-runSkill
- challenge-handoffSkill
- dot-diagramsSkill
- handoffSkill
- implem-brakeSkill
- resume-handoffSkill
- review-auto-loopSkill
- review-correctnessSkill
- review-docsSkill
- review-judgeSkill
- review-readabilitySkill
- review-testsSkill
- rust-project-maintenanceSkill
- settle-discussionSkill
- web-searchSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
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 30
Never work from a partial diff: read all of it first, fetching whatever a truncated or elided output left out.
- Distinctive: line 32
Read the surrounding code wherever the diff alone does not say what the changed code does.
- Distinctive: line 41
When a feature or a bug fix comes with its tests or its documentation, they earn at most a trailing clause.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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
Summarize Jujutsu changes
Describe what the Jujutsu changes for the target revision do, in prose.
Input
A summary targets a single Jujutsu revision, <JJ_REVISION>, supplied by the user. Take it verbatim from the prompt.
Resolve it yourself only when the prompt carries none:
jj log -r 'latest(::@ & ~empty())' --no-graph -T 'change_id'Whether given or resolved, <JJ_REVISION> is fixed for the whole summary: use that exact value everywhere below.
Run:
jj show <JJ_REVISION>Never work from a partial diff: read all of it first, fetching whatever a truncated or elided output left out.
Read the surrounding code wherever the diff alone does not say what the changed code does.
Main goals
Describe the change from the diff. The revision description says what its author meant to do; the summary says what the code does, and never restates the description.
- Describe, never interpret: say what the code now does, not why it was written, what it enables, or what it improves
- Stay high level: no code, no pseudo code, no file by file walkthrough, no line references
- Give each unrelated change exactly one paragraph, ordered by decreasing weight in the diff
- When a feature or a bug fix comes with its tests or its documentation, they earn at most a trailing clause; test or documentation work beyond them, like rewriting existing sections, is a change of its own, described as the subject it is
Output
Your final message is the summary and nothing else, with no preamble and no recap.
One or more paragraphs, and nothing else: no heading, no list, no code block, no blockquote, no table. A code span naming a symbol the change touches is fine.
Two to five sentences per paragraph, and under 200 words in total.
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.

