setup
Skill39 starsChanged 22 days agoMITView on GitHub
Interview the user one question at a time and write the project's rules file, goals file, handoff file and release checklist. Use when starting a new app, adopting The Last Mile in an existing project, or when the user asks to set up or redo their project rules.
191 other projects use a version of this.
Adopted by
Most recently taken up by
- hydra-db/hydradb-claude-codetheir own version, first seen Nov 8, 2026
- WALKERKILLER/Look-Tongji-Notestheir own version, first seen Nov 8, 2026
- steerlabs/opensteertheir own version, first seen Nov 8, 2026
- ansvisor/ansvisortheir own version, first seen Nov 8, 2026
- dsifry/metaswarmtheir own version, first seen Nov 8, 2026
- anombyte93/prd-taskmastertheir own version, first seen Nov 8, 2026
Used with
Projects with it also use add-voice-transcription, add-discord and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
juancruzmunozalbeloSkilladd-voice-transcription
Add voice message transcription to NanoClaw using OpenAI's Whisper API. Automatically transcribes WhatsApp voice notes so the agent can read and respond to them.
Also in 5 projects that use this
MicrosoftSkilladd-discord
Add Discord bot channel integration via Chat SDK.
Also in 11 projects that use this
MicrosoftSkilladd-slack
Add Slack channel integration via Chat SDK.
Also in 11 projects that use this
juancruzmunozalbeloSkilladd-telegram-swarm
Add Agent Swarm (Teams) support to Telegram. Each subagent gets its own bot identity in the group. Requires Telegram channel to be set up first (use /add-telegram). Triggers on "agent swarm", "agent teams telegram", "telegram swarm", "bot pool".
Also in 7 projects that use this
- +8 more from microsoft
Other versions191 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 191
- anthropics/knowledge-work-plugins★ 35ksame name
- garrytan/gbrain★ 32ksame name
- gavrielc/nanoclaw★ 31ksame name
- nanocoai/nanoclaw★ 31ksame name
- qwibitai/nanoclaw★ 31ksame name
- Q00/ouroboros★ 6.3ksame name
- pytorch/executorch★ 5.2ksame name
- agenticnotetaking/arscontexta★ 3.5ksame name
- aws/agent-toolkit-for-aws★ 3.4ksame name
- spotify/portal-ai-plugins★ 2.8ksame name
181 more not listed
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 repository21 other files it gives its agents.
- .claude/CLAUDE.mdCLAUDE.md
- build-a-baselineSkill
- connect-toolsSkill
- extract-lessonsSkill
- fact-checkSkill
- feature-demoSkill
- improve-skillSkill
- lean-outSkill
- mine-communitySkill
- one-time-surfaceSkill
- port-featureSkill
- reference-huntSkill
- researchSkill
- screen-atlasSkill
- search-pagesSkill
- see-it-workSkill
- session-endSkill
- set-goalsSkill
- shipSkill
- store-listingSkill
- syncSkill
The file
We split this skill into the rules it gives an agent
8 rules. 0 are common, 5 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 170 edited versions we read keep these.
5
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
- Very specific: line 8
Run a short interview, then write four files from the templates in this skill's templates/ folder.
- Distinctive: line 13
One question at a time: Wait for the answer before asking the next.
- Very specific: line 15
Read before asking: If the project already has a CLAUDE.md, AGENTS.md, README or app.json, read them first and only ask what they do not …
- Distinctive: line 29
Ask these in order. Skip any the project already answers.
- Distinctive: line 31
What does the app do?: Ask for one sentence.
- Distinctive: line 40
Default: a branch with a pull request, unless they work alone and prefer main.
- Very specific: line 41
What command runs your checks?: Default: whatever package.json defines as check, or lint, typecheck and test in that order.
- Distinctive: line 43
Run the command once before moving on.
“Common” is measured: kept by at least 6 in 10 of the 170 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
Setup
Run a short interview, then write four files from the templates in this skill's
templates/ folder. The whole thing should take about 10 minutes of the user's time.
Rules for the interview
- One question at a time. Wait for the answer before asking the next.
- Offer a sensible default in every question so a short answer is enough.
- Read before asking. If the project already has a CLAUDE.md, AGENTS.md, README or app.json, read them first and only ask what they do not already answer. Confirm what you inferred in one line instead of asking it again.
- Look at the user's other projects too. Sibling folders (the project's parent, and wherever their other repos live) may hold a rules file that already has a How to talk to me section or standing decisions marked as applying to every repo. Offer to carry those over in one line instead of asking question 4 and the push question again. Copy only the personal parts, never another app's settings.
- Never overwrite an existing file without showing the user what would change and getting a yes. Prefer merging new sections into an existing rules file.
- A screenshot is a perfectly good answer. If the user sends one, use it.
The questions
Ask these in order. Skip any the project already answers.
- What does the app do? Ask for one sentence.
- What is the loop? The thing a person does, start to finish, that gives them what they came for. Example: "add a plant, get a reminder, mark it watered". If they cannot say it in one sentence, help them cut it down.
- Who is it for? One specific kind of person and the moment they reach for the app.
- How should Claude talk to you? Offer choices: short and direct; detailed with reasoning; plain words with no code names. Ask if there is anything they never want to see (some people dislike long dashes, jargon, or recaps).
- How do you push code? Straight to main, or to a branch with a pull request. Default: a branch with a pull request, unless they work alone and prefer main.
- What command runs your checks? Default: whatever
package.jsondefines ascheck, or lint, typecheck and test in that order. Offer to add acheckscript. Run the command once before moving on. A first run can install tools or write config files (Expo's linter does both); those changes belong to setup, and the user should know whether the checks start clean. - What is your first 6 week goal? Offer the default for their stage:
- Idea only: "A working version of the loop on my own phone by
<date>." - Working build: "Live in the App Store (and Google Play) by
<date>." For a website: "Live at its own domain, with real people signed up, by<date>." - Live: "First 100 people using it every week by
<date>." Then ask for the one number that shows it is working, and 3 to 5 must dos. The set-goals skill'sreferences/stages.mdhas fuller defaults for each stage.
- Idea only: "A working version of the loop on my own phone by
- Where does it run? Both stores, iOS only, Android only, or web only (a website that never goes to a store). Default: both stores, unless the code is a web framework with no app build, then web only. Mention that a new personal Google Play account must run a 14 day closed test first, which is why some people ship iOS first.
- How much before the first launch? Offer the three levels from the ship skill's
references/levels.md, one line each: Quick ship (the least the stores accept, free at launch), Standard (the default: crash reporting, a tested release build, a listing written to be found, paid features if any), Full features (analytics, subscriptions with a trial, preview video, localized listings). Skip this question for a web only project and writenot applicable (web).
What to write
| File | From template | Notes |
CLAUDE.md | templates/rules-file.md | If the user's agent reads AGENTS.md instead, write that and make CLAUDE.md point to it |
GOALS.md | templates/goals.md | End date is 6 weeks from today |
CURRENT_WORK.md | templates/current-work.md | The handoff file session end keeps current |
RELEASE_TEST_CHECKLIST.md | templates/release-checklist.md | For a web only project, items are checked on the live site after each deploy |
Fill every placeholder in angle brackets. Date every entry you write, in
YYYY-MM-DD form. Remove template sections that do not apply rather than leaving
them empty.
Finish
Show the user a short summary: the four files written, their goal and end date, and the skills to use next: connect-tools now, to connect the services the app uses, session-end every session, see-it-work to see a change working, reference-hunt before designing a screen, research before a decision, set-goals when the cycle ends, ship when getting into the stores, and setup again if anything about the answers changes. Then offer to commit the files, following the push choice from question 5.
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.

