simple-plan
Skill31 starsChanged 6 months agoMITView on GitHub
Lightweight planning for small, quick changes. Combines propose, spec, and (optionally) design into one conversational pass, then writes tasks.md so the change is ready for implementation. Activates when users ask for "simple plan", "quick plan", "plan this", or "planning mode" for a change that doesn't warrant the full propose/spec/design ceremony.
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 3 are distinctive, and 3 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.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 20
Use codagent:ask-questions to resolve the decisions behind the proposal, specs, and any design:
- Distinctive: line 34
Write all artifacts in one pass, then check them together for missing decisions or contradictions.
- Very specific: line 37
Follow a project-defined location.
- Distinctive: line 45
Keep the proposal brief: why, high-level changes, new and modified capabilities, out of scope, and impact.
- Very specific: line 78
Use ## REMOVED Requirements with Reason and Migration, and ## RENAMED Requirements with FROM and TO, when those delta operations apply.
- Distinctive: line 99
Write one placeholder entry linking every artifact:
≈ “Common” is an estimate: at least 1 in 5 of the 308,232 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
Simple Plan
Plan a small change in one focused conversation. Produce a concise proposal.md, one spec per
capability, an optional design.md, and a tasks.md handoff. The artifacts must be self-contained for
an implementing agent with no conversation history.
Process
- Understand the problem, desired behavior, success criteria, and scope. Lightly inspect related specifications and relevant code so questions and artifacts reflect the existing system.
Use
codagent:ask-questionsto resolve the decisions behind the proposal, specs, and any design:- Proposal: the problem, desired outcome, success criteria, scope boundaries, and any choice between real alternatives that changes product scope or direction.
- Specs: observable behavior, boundaries, errors, and edge cases that affect behavior or scope.
- Design: consequential architectural choices that repository context cannot safely resolve.
Recommend a default when useful, decide low-risk implementation details from repository context, and do not use a generic approval question as discovery.
- Decide whether
design.mdis necessary. Default to no; write one only for a consequential architectural choice, non-obvious rationale, migration, integration strategy, or implementation constraint that the implementer otherwise would not know. - Present the proposed capability set, key requirements, design-doc decision, output location, and the consequential assumptions or defaults you selected, so the user can correct them. Wait for approval before writing.
- Write all artifacts in one pass, then check them together for missing decisions or contradictions. Return to the user for any new behavioral or scope decision instead of inventing one.
Follow a project-defined location. Otherwise propose ~/.agent-skills/changes/<kebab-slug>/ at the
approval step. Do not run separate proposal, spec, and design phases, and do not write a detailed task
breakdown.
Artifact requirements
proposal.md
Keep the proposal brief: why, high-level changes, new and modified capabilities, out of scope, and impact.
## Why
<problem or opportunity>
## What Changes
- <high-level change>
## Capabilities
### New Capabilities
- `<name>`: <brief description>
### Modified Capabilities
- `<existing-name>`: <changed requirements>
## Out of Scope
<explicit exclusions>
## Impact
<affected code, APIs, dependencies, or users>specs/<capability>/spec.md
Specifications are the load-bearing artifact. Use SHALL or MUST, observable behavior, and at least one
#### Scenario: with WHEN/THEN per requirement. For a modified requirement, copy the complete existing
requirement and all scenarios under ## MODIFIED Requirements before editing it. If behavior truly
depends on an unresolved architectural choice, mark the scenario
<!-- deferred-to-design: <reason> --> and resolve it in the design.
Use ## REMOVED Requirements with Reason and Migration, and ## RENAMED Requirements with
FROM and TO, when those delta operations apply.
## ADDED Requirements
### Requirement: <name>
<normative behavior>
#### Scenario: <name>
- **WHEN** <condition>
- **THEN** <observable result>design.md when needed
Record only the context, approach, decisions and rationale, risks or trade-offs, and migration details needed for implementation.
tasks.md
Write one placeholder entry linking every artifact:
- [ ] Implement the change described by these files:
- [proposal.md](proposal.md)
- [specs/<capability>/spec.md](specs/<capability>/spec.md)
- [design.md](design.md) <!-- only when written -->In this repository30 other files it gives its agents.
- AGENTS.mdAGENTS.md
- docs/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- releaseSkill
- releaseSkill
- ask-questionsSkill
- call-agentSkill
- designSkill
- finalize-prSkill
- fix-prSkill
- handoffSkill
- implement-changeSkill
- initSkill
- orchestrate-changeSkill
- plan-tasksSkill
- prepare-acceptanceSkill
- proposal-reviewSkill
- proposeSkill
- push-prSkill
- review-approachSkill
- review-assumptionsSkill
- review-specSkill
- review-tasksSkill
- session-reportSkill
- specSkill
- task-complianceSkill
- test-flowsSkill
- test-planSkill
- verify-behaviorSkill
- wait-ciSkill
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.

