Generates a detailed implementation plan by researching the codebase and reconciling project intent.
Install
mkdir -p .claude/skills/plan-dcouple && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/17484" && unzip -o skill.zip -d .claude/skills/plan-dcouple && rm skill.zipInstalls to .claude/skills/plan-dcouple
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Creates a reconciled implementation plan by combining a structured plan draft with a normalized intent brief and a PRP-style research dossier, then auto-reviews the final plan. Use when planning a new feature or significant change in Codex.Key capabilities
- →Verify current repository shape for feature area
- →Clarify requirements by asking targeted questions
- →Conduct external research (library docs, best practices)
- →Draft a provisional implementation plan
- →Create a normalized brief/intent artifact
- →Generate a supporting research dossier
How it works
The skill performs a repo audit, clarifies requirements, conducts external research, then drafts a provisional plan, a normalized brief, and a research dossier, which are then reconciled into a final plan.
Inputs & outputs
When to use plan
- →Plan new features
- →Research implementation path
- →Reconcile project architecture
About this skill
Plan
Generate a complete plan for feature implementation with thorough research. The plan must contain enough context for an AI agent to implement the feature in a single pass.
Codex is the primary planner in this workflow. If you also have a separate Claude workflow available, treat it as an optional second-opinion lane rather than the source of truth.
Step 1: Mandatory Repo Audit
Do not start drafting until you have verified the current repo shape for the feature area.
Verify These Facts In-Repo
- Primary entrypoint(s) and integration surfaces relevant to this feature
- Exact module names and singular/plural usage
- Validator/controller/service directory layout in the affected area
- Actual data-model/schema/type source of truth used by this codebase
- Existing user-facing or operator-facing surface(s) this feature extends
- Shared type/export hubs if cross-app types are needed
- Actual validation/build/typecheck workflow used by this repo
Repo Audit Rules
- Do not assume any specific stack or layout. Discover the actual routing, validation, schema, frontend, and build patterns used by the current repo.
- Every existing file path cited in the final plan must have been opened in this session.
- Mark every path in the final plan as either
existingornew. - Never cite a line number unless it was verified in the current checkout.
- Never let template/example paths leak into the final plan.
- If the brief or user request conflicts with repo reality, add a
Known Mismatches / Assumptionssection that states the conflict and how the plan resolves it.
Step 1b: Clarify Requirements (Only If Needed)
If, after the repo audit, the approach is genuinely unclear, ask the user 1-3 targeted design questions. Otherwise, proceed directly.
Step 1c: External Research (Only If Needed)
- Library documentation
- Implementation examples
- Best practices and common pitfalls
- Prefer primary documentation when researching external behavior
Step 2: Draft the Plan, Intent Artifact, and Research Dossier
Produce three artifacts from the same brief:
- A provisional implementation plan using
./plan_base.md - A normalized brief / intent artifact that preserves the why, locked decisions, non-goals, and success criteria in a compact downstream-friendly form
- A supporting research dossier that behaves like a PRP: anchor-dense, selective, and focused on context transfer
The final output shown to the user is the reconciled plan, not the dossier.
Step 2a: Draft the Provisional Plan
Use ./plan_base.md in this skill directory as the template.
Critical Context to Include
The AI agent only gets the context in the plan plus codebase access. Include:
- Intent / Why
- Verified Repo Truths
- Evidence with exact
file:line-line - Locked Decisions
- Documentation URLs when needed
- Code Examples from the codebase
- Gotchas
- Patterns to follow
- Known Mismatches / Assumptions
- Critical Codebase Anchors
Plan Guidelines
- Required sections are: Summary, Intent / Why, Source Artifacts, Verified Repo Truths, Locked Decisions, Known Mismatches / Assumptions, Critical Codebase Anchors, Files Being Changed, Reconciliation Notes, Delta Design, Architecture Overview, Key Pseudocode, Tasks, Validation, and Open Questions.
Verified Repo Truthscontains facts only.- Every fact needs
Fact,Evidence, andImplication. - Negative or absence-based claims also need
Search Evidence. - If it is not proven, it is not a fact.
- Every
MODIFYpath must already exist. - Do not leak placeholder/example paths into the final plan.
- Keep repo facts separate from proposed changes.
- Mirror current codebase patterns rather than inventing approximate examples.
- Do not add compatibility layers unless the user explicitly asks.
- Do not add unit or integration tests by default.
- Use
[NEEDS CLARIFICATION]markers instead of guessing.
Step 2b: Create a Normalized Brief / Intent Artifact
Save a normalized brief / intent artifact at:
./tmp/plan-artifacts/YYYY-MM-DD-description-brief.md
This is a compact intent capsule for downstream implementation and review. Include:
- Problem / outcome summary
- Who this matters for
- Locked decisions already made
- Non-goals / what must not be optimized away
- Success criteria
- Explicit user constraints
The final plan must record this path in Source Artifacts.
Step 2c: Create a Research Dossier
Save a supporting dossier at:
./tmp/plan-artifacts/YYYY-MM-DD-description-research-dossier.md
The dossier should:
- behave like a PRP-style supporting artifact, not the final plan
- focus on critical codebase anchors, patterns to reuse, gotchas, external docs, and a suggested implementation shape
- use exact
file:line-linereferences for repo claims - include external docs only when they materially reduce risk
- avoid placeholder text and generic examples
Step 3: Reconcile the Dossier into the Final Plan
Before saving the user-facing plan, compare the provisional plan against the research dossier and reconcile them.
Reconciliation Goals
- Import missing anchors from the dossier into the final plan
- Import missing docs, gotchas, and load-bearing constraints
- Preserve the brief's why, locked decisions, and non-goals as first-class constraints in the final plan
- Surface factual conflicts between the draft and dossier
- Remove duplicated or low-value sections
- Preserve a clean separation between verified facts, settled decisions, and proposed changes
Reconciliation Rules
- The final plan is authoritative
- The brief / intent artifact is authoritative for why
- Do not paste the dossier wholesale into the plan
- If the plan and dossier disagree, re-check the repo before choosing a side
- If a simplification weakens the brief's intent, surface it rather than hide it
- Do not import unsupported dossier claims into
Verified Repo Truths - Keep only the highest-value anchors, patterns, docs, and gotchas
- Add concise
Reconciliation Notes
Pre-Save Reality Check
Before saving the plan, verify all of the following:
- Every
MODIFYpath exists - No placeholder/example paths remain
- Every line anchor was checked in the current checkout
- Every
Verified Repo Truthsbullet includesFact,Evidence, andImplication - Every negative claim includes
Search Evidence - No future/proposal language appears inside
Verified Repo Truths - Entry points and integration points match the repo audit
- Code examples match current helper patterns
- The dossier has been compared against the provisional plan
- Any plan-vs-dossier conflicts were resolved or surfaced explicitly
Step 4: Save the Final Plan and Supporting Artifacts
Save the final reconciled plan as:
./tmp/ready-plans/YYYY-MM-DD-description.md
Save the supporting research dossier as:
./tmp/plan-artifacts/YYYY-MM-DD-description-research-dossier.md
Save the normalized brief / intent artifact as:
./tmp/plan-artifacts/YYYY-MM-DD-description-brief.md
Only the reconciled plan belongs in ready-plans.
Step 5: Review and Present
After saving the plan, run the review gates.
- Run a skeptical review against the standards in
plan-reviewer. - If you can run a fresh second review context, do it and compare results.
- If you are operating alongside a separate Claude workflow, you may use that as the parallel second-opinion lane, but Codex remains the primary planner.
- Split findings into:
- Auto-fixable
- Needs user input
- Apply all auto-fixable changes silently.
- Do not surface questions until all active review lanes are complete and their findings are merged.
Present to the User
- Plan Summary: 3-5 bullets
- Questions for You: only genuine decisions or unresolved ambiguity
- Plan Link:
./tmp/ready-plans/[filename] - Optional links:
- Brief / intent artifact
- Research dossier
- End with:
Want to run another review pass, or is this ready to implement?
If the user wants changes or another review pass, apply the changes and rerun a fresh review.
Do not treat the plan as ready if factual blockers remain unresolved.
Step 6: Return the Plan — Do Not Implement
Once the user confirms the plan is ready, tell them:
Plan finalized! To implement, run:
/implement ./tmp/ready-plans/[filename]
Your job ends here. Do not start implementing the plan in the same step.
Quality Checklist
- Supporting research dossier created
- Supporting brief / intent artifact created
- Existing file paths verified in-session
- No placeholder/example paths leaked from the template
- Plan includes
Intent / WhyandSource Artifacts - Verified Repo Truths contains facts only
- Every verified fact has exact evidence
- Every negative claim has search evidence
- High-value anchors/docs/gotchas from the dossier were reconciled into the plan or intentionally dropped
- The brief's why, locked decisions, and non-goals survived reconciliation
- Validation gates are executable by AI
- References existing patterns
- Clear implementation path
- Error handling documented
When not to use it
- →When the user wants autonomous end-to-end execution
- →When the user wants to start coding immediately with a clear task
- →When the task is a single focused fix with obvious scope
Limitations
- →Does not implement the plan
- →Does not add unit or integration tests by default
- →Does not add compatibility layers unless explicitly asked
How it compares
This workflow generates a complete, research-backed implementation plan before any coding, ensuring a solid foundation, unlike directly jumping into code.
Compared to similar skills
plan side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| plan (this skill) | 0 | 3mo | No flags | Advanced |
| openspec-explore | 4 | 5mo | Review | Beginner |
| thought-patterns | 6 | 7mo | No flags | Advanced |
| gathering-requirements | 3 | 8mo | No flags | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by dcouple
View all by dcouple →You might also like
openspec-explore
studyzy
进入探索模式 - 一个用于探索想法、调查问题和澄清需求的思考伙伴。当用户想要在进行更改之前或期间深入思考某事时使用。
thought-patterns
markpitt
Orchestrates comprehensive cognitive thinking patterns including sequential, systems, lateral, divergent/convergent, abductive, analogical, first-principles, double-loop learning, gestalt, metacognitive, and neurodivergent patterns (ADHD hyperfocus, autistic detail-orientation, dyslexic spatial reasoning). Analyzes tasks to select optimal pattern(s), chains multiple patterns when needed, and validates outputs before responding. Use for complex problem-solving, creative tasks, analytical challenges, or when diverse cognitive approaches enhance solution quality.
gathering-requirements
CaptainCrouton89
Systematically clarify user needs, preferences, and constraints before planning or implementation. Classifies work type, investigates existing systems, discovers edge cases and integration points, resolves assumptions, and creates detailed specifications. Use when building features, enhancements, or integrations where requirements need clarification.
the-fool
Jeffallan
Use when challenging ideas, plans, decisions, or proposals using structured critical reasoning. Invoke to play devil's advocate, run a pre-mortem, red team, or audit evidence and assumptions.
decision-critic
solatis
Invoke IMMEDIATELY via python script to stress-test decisions and reasoning. Do NOT analyze first - the script orchestrates the critique workflow.
tech-divergence
WellApp-ai
Evaluate technical options with scoring matrix, trigger Gate 4 for significant decisions