one-by-one
Skill4 starsChanged 6 days agoMITView on GitHub
Work through a list of items (review comments, findings, report points, open questions) one at a time. For each item, give context, a preview of the change, and an opinion, then reach agreement with the user before applying anything. Use when the user says "one by one", "iteratively", "discuss each point", or asks to go through a list of changes together.
5 other projects use a version of this.
Adopted by
Most recently taken up by
- p3bot/librarytheir own version, first seen Oct 31, 2026
- rupertsworld/skillstheir own version, first seen Oct 13, 2026
- kjendrzyca/public-skillstheir own version, first seen Oct 8, 2026
- BackToCimaCoppi/Praxistheir own version, first seen Sep 23, 2026
- YoniChechik/claude-code-configtheir own version, first seen Sep 18, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- p3bot/library★ 10same name
- BackToCimaCoppi/Praxis★ 6same name
- kjendrzyca/public-skills★ 1same name
- rupertsworld/skills★ 1same name
- YoniChechik/claude-code-config★ 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 repositoryOne other file it gives its agents.
The file
We split this skill into the rules it gives an agent
10 rules. 0 are common, 10 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
10
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 16
If the user names no source (for example, they only invoke the skill), use your own latest message in the conversation as the source.
- Distinctive: line 24
Do not start on an item yet.
- Distinctive: line 29
Use this structure for each item:
- Distinctive: line 43
Keep each turn short. Do not present the next item in the same turn.
- Distinctive: line 50
If the change causes a new question, do not decide it yourself.
- Distinctive: line 54
If the user brings new information (for example a conversation with a colleague), re-evaluate the item with it.
- Distinctive: line 60
Do not commit, push, post replies, resolve threads, or edit pull requests or issues unless the user asks for that in this conversation.
- Distinctive: line 69
After the last item:
- Distinctive: line 74
Run the project checks that apply to the changed files, and report the result.
- Distinctive: line 76
Ask whether to commit, and suggest a commit message.
“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
One-by-one consensus
Work through a list of items with the user, one item per turn. The user decides each item. You apply only what the user approves.
1. Collect and list
Collect the items from the source the user names: review comments, a report, findings, or a document.
If the user names no source (for example, they only invoke the skill), use your own latest message in the conversation as the source. Take the items, findings, or proposals from it. If that message contains no items to decide, ask the user what to work through.
Show a numbered list with a one-line summary of each item. Group related items when one change addresses both, and say so.
Do not start on an item yet. Suggest an order (largest impact first, or dependencies first) and start when the user agrees.
2. Present one item per turn
Use this structure for each item:
- Context: quote the source (the comment, or the current text with line numbers). Name the earlier decisions that the item depends on. Introduce any concept that the user may not know.
- Preview: show the exact change as a diff or as the replacement text. Do not edit files yet.
- My opinion: give a recommendation and the reason. Push back when you disagree, also with the reviewer or the user. When there is a real choice, give labelled options (A, B, C) so the user can answer in one word. Name what the change loses, not only what it gains.
- End with one question: apply, change, or skip? If there are options, ask which one.
Keep each turn short. Do not present the next item in the same turn.
3. Apply only what the user approves
- On approval, apply exactly the approved variant.
- In the same edit, fix direct knock-on references (for example a sentence elsewhere that now contradicts the change). Name each one in one line.
- If the change causes a new question, do not decide it yourself. Raise it as a note, or as its own item.
- On "skip", change nothing and move on.
- On "change", show a new preview before you apply it.
- If the user brings new information (for example a conversation with a colleague), re-evaluate the item with it. Push back if it has a cost.
- Confirm each applied change in one line, then present the next item.
4. Do not act outside the agreement
- Do not commit, push, post replies, resolve threads, or edit pull requests or issues unless the user asks for that in this conversation.
- Approval of one action does not extend to the next one. Each commit, push, or external message needs its own request.
- Draft replies and messages in the chat only, unless the user asks you to send them.
5. Close
After the last item:
- Summarise what you applied and what the user skipped, for example in a short table.
- List the open points that remain.
- Run the project checks that apply to the changed files, and report the result.
- Ask whether to commit, and suggest a commit message.
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.