plan-in-nine-steps
Skill1 starsChanged 4 days agoMITView on GitHub
Plan a feature in nine steps using the checklists in steps/, starting with a five-line executive summary and ending with the delivery backlog. Loads only the brief and approved in-scope brain entries, drafts one step at a time, and stops for a person to approve each. Use when someone says "plan this feature", "plan in nine steps" or "draft the next step".
In this repository5 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- brain-healthSkill
- captureSkill
- write-handoffSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 3 are distinctive, and 4 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.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 12
Ask which feature (features/<slug>/).
- Very specific: line 13
Read AGENTS.md, features/<slug>/brief.md and brain/INDEX.md.
- Distinctive: line 16
Check that every earlier step this one reads has been approved in the brief's steps table.
- Very specific: line 18
Write the step to features/<slug>/<step>.md.
- Very specific: line 25
When answers arrive, run the capture skill so the answers worth keeping are saved with a scope.
- Distinctive: line 29
Start the next step before the person approves this one.
- Distinctive: line 30
Treat a draft step or proposed entry as fact.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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
Plan in nine steps
One step at a time. Each step reads only approved earlier steps. Start with a five-line summary, and rewrite it at the end.
How to run it
- Ask which feature (
features/<slug>/). If it does not exist, copyfeatures/_template/and ask the person to fill inbrief.md. - Read
AGENTS.md,features/<slug>/brief.mdandbrain/INDEX.md. Work out the scope from the brief. - Open only entries that are
status: approvedand whose scope matches. Read rules with severitymustfirst. Ignore proposed entries. - Pick the next step in this order: 01 executive summary (first pass), 02 intent, 03 impact, 04 UX, 05 backend, 06 system impact, 07 data, 08 plan, 09 backlog, then 01 executive summary again (final pass). Skip a step marked not needed.
- Check that every earlier step this one reads has been approved in the brief's steps table. If not, stop and say which one is missing.
- Open
steps/<step>.mdand work through its checklist. Read the approved earlier steps it names. Do not read drafts. - Write the step to
features/<slug>/<step>.md. In it:- Cite entry IDs for everything you rely on.
- Label each assumption with a confidence and what would disprove it.
- List sources.
- Say "to be confirmed" instead of guessing.
- List questions in
features/<slug>/questions.mdwith an owner role and a blocking level. If a blocking question is open, stop here and ask. - Set the step to
draftin the brief and ask the person to review and approve it. Only a person changes it toapprovedand fills in approver and date. - When answers arrive, run the
captureskill so the answers worth keeping are saved with a scope.
Do not
- Start the next step before the person approves this one.
- Treat a draft step or proposed entry as fact.
- Invent file paths, module names or endpoints.
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.

