design-before-plan
Skill2 starsChanged 36 days agoMITView on GitHub
Use before creative or behavior-changing work — tease out approaches and a short design for Captain sign-off before writing IMPLEMENTATION_PLAN.md
In this repository57 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .cursor/rules/00-core-operating-model.mdcCursor rule
- .cursor/rules/01-plan-approval-gate.mdcCursor rule
- .cursor/rules/02-git-worktree-policy.mdcCursor rule
- .cursor/rules/03-validation-definition-of-done.mdcCursor rule
- .cursor/rules/04-documentation-memory.mdcCursor rule
- accessibility-reviewSkill
- agent-skills-packagingSkill
- autonomy-budgetSkill
- bounded-autonomySkill
- candidate-promotionSkill
- capability-planningSkill
- code-reviewerSkill
- code-structure-cleanupSkill
- compass-evaluatorSkill
- context-inodesSkill
- dependency-supply-chainSkill
- docker-cloudSkill
- embedding-providersSkill
- execution-telemetrySkill
- experience-routingSkill
- experience-skill-trainingSkill
- external-knowledge-ingestSkill
- finishing-a-development-branchSkill
- github-integrationSkill
- goose-agent-orchestrationSkill
- harness-gcSkill
- hosted-vector-dbSkill
- implementation-planningSkill
- ios-engineeringSkill
- knowledge-stewardSkill
- linear-integrationSkill
- node-engineeringSkill
- northstar-connected-routineSkill
- notion-integrationSkill
- package-registry-tiSkill
- persistent-role-promotionSkill
- playwright-browser-validationSkill
- postgres-prismaSkill
- post-merge-wave-continueSkill
- procedure-playbooksSkill
- project-overseerSkill
- pull-request-preparationSkill
- python-mlSkill
- react-engineeringSkill
- repository-discoverySkill
- review-fix-loopSkill
- security-audit-proceduresSkill
- security-reviewSkill
- skill-learning-loopSkill
- skill-lifecycleSkill
- source-code-contextSkill
- subagent-driven-developmentSkill
- technology-intelligence-liveSkill
- testing-validationSkill
- verification-before-completionSkill
- worktree-orchestrationSkill
The file
We split this skill into the rules it gives an agent
10 rules. About 0 are common, 7 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
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 15
Before /plan-feature / writing IMPLEMENTATION_PLAN.md
- Distinctive: line 19
When the request already specifies everything that matters (e.g.
- Distinctive: line 31
Look before you ask: Read referred files and related code until you can explain how things work today and what would change.
- Distinctive: line 35
Ask the Captain to pick one (or amend).
- Very specific: line 38
Get explicit design approval: before creating or updating IMPLEMENTATION_PLAN.md.
- Very specific: line 40
After design approval, load Skill implementation-planning / run /plan-feature and produce the approval-gated plan as usual.
- Distinctive: line 53
Do not skip the Captain plan-approval gate after design
- Distinctive: line 54
Do not implement product files during design
- Distinctive: line 55
Do not install third-party Superpowers/ECC plugins
- Distinctive: line 56
Do not reopen decided Captain choices without new evidence
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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
Design Before Plan
Patterns inspired by obra/superpowers
brainstorming (DISCOVERED inspiration — not a plugin install).
Use this Skill when
- Creating features, changing product behavior, or non-trivial design choices
- The request leaves approach, UX, or architecture open
- Before
/plan-feature/ writingIMPLEMENTATION_PLAN.md
Skip (quick tasks)
When the request already specifies everything that matters (e.g. “rename X to Y”, “make the icon cornflower blue”), do the work under normal rules — do not force a design ceremony.
Inputs
- Captain request
- Existing project memory and relevant code/docs
- Any prior decisions the Captain already made
Procedure
- Look before you ask. Read referred files and related code until you can explain how things work today and what would change. Do not re-ask what the repo or Captain already answered.
- Lay out approaches in their own message (2–3 realistic options, trade-offs). Ask the Captain to pick one (or amend).
- Write a short design (or fuller design doc for large work): goal, chosen approach, boundaries, non-goals, risks. Size to what a Captain can actually read.
- Get explicit design approval before creating or updating
IMPLEMENTATION_PLAN.md. - After design approval, load Skill
implementation-planning/ run/plan-featureand produce the approval-gated plan as usual.
Spikes the Captain explicitly approved are exploration only — still no product behavior ship without a later plan gate when implementation follows.
Output
- Approaches + chosen design (Captain-approved)
- Handoff into
IMPLEMENTATION_PLAN.md(still AWAITING APPROVAL until plan gate)
Prohibited actions
- Do not skip the Captain plan-approval gate after design
- Do not implement product files during design
- Do not install third-party Superpowers/ECC plugins
- Do not reopen decided Captain choices without new evidence
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.

