agentsclimarketplace

Deep plan

Skill josephneumann/claude-corps/skills/deep-plan

Use when starting a new feature or idea that needs thorough pre-execution planning. Use instead of calling /product-review, /spec, and review skills individually.From its SKILL.md

Install
npx -y skills add josephneumann/claude-corps --skill deep-plan

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

  • skips confirmationTells the agent to proceed without asking first, 3 times: "If `--yes`: proceed automatically." and 2 more.
  • 4 stars4 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
  • runs commandsInstructs the agent to run 2 commands, including `ls -t docs/plans/*.md 2>/dev/null | head -1` and 1 more.

SKILL.md

6.4 KB, ~1.5k tokens by cl100k_base, as published. Nobody here has run it

/deep-plan $ARGUMENTS

Orchestrate the full planning pipeline in the correct order. Each phase invokes an existing skill via the Skill tool and presents a checkpoint before the next phase.

This skill sequences sub-skills. It does NOT duplicate their content.

Argument Parsing

Arguments: $ARGUMENTS

Parse flags:

  • --yes — Auto-answer all checkpoints with defaults (opt-in phases skip, opt-out phases run). Pass --yes to sub-skills that support it.
  • --skip-reviews — Skip Phases 4-6 (deep eng + design + multi-review). Use when spec's Phase 2.5 inline reviews are sufficient.
  • --no-decompose — Skip Phase 7 (Linear decomposition).
  • Everything else = the feature description. Store as $FEATURE.

If $FEATURE is empty, ask: "What feature or idea do you want to plan?"

Do not proceed without a feature description.

Phase 1: Product Review (always)

Run /product-review via Skill tool with $FEATURE as arguments.

Wait for completion. The sub-skill handles its own internal checkpoints (interrogation, DVF gate, mode selection).

Checkpoint:

If --yes: proceed automatically.

Otherwise, use AskUserQuestion:

  • Question: "Product review complete. Continue to planning?"
  • Options: "Continue to planning" (recommended) / "Stop here — need to rethink"

If user stops: output "Pipeline stopped after product review. Re-run /deep-plan when ready." Then STOP.

Phase 2: Design Exploration (opt-in)

If --yes: skip this phase (default = skip for opt-in).

Otherwise, use AskUserQuestion:

  • Question: "Explore 3-5 different UI directions before writing the spec? Recommended for UI-heavy features."
  • Options: "Yes, explore directions" / "No, go straight to spec" (recommended)

If yes: Run /design-shotgun via Skill tool with $FEATURE as arguments. Wait for completion.

If no: Proceed.

Phase 3: Spec (always)

Run /spec via Skill tool with $FEATURE as arguments.

/spec runs Phase 2.5 (lightweight eng + design review) inline and writes a plan file to docs/plans/. Wait for completion.

After /spec finishes, identify the plan file path. Store as $PLAN_PATH:

ls -t docs/plans/*.md 2>/dev/null | head -1

Checkpoint:

If --yes: proceed to Phase 4 (unless --skip-reviews, then jump to Phase 7).

Otherwise, use AskUserQuestion:

  • Question: "Spec written to $PLAN_PATH. Phase 2.5 ran inline eng + design checks. What next?"
  • Options: "Continue to deep reviews" (recommended) / "Skip reviews, go to decompose" / "Stop here — spec is the deliverable"

If "Skip reviews": jump to Phase 7. If "Stop here": jump to Pipeline Summary.

Phase 4: Engineering Review (opt-out)

If --skip-reviews: skip to Phase 7.

If --yes: run automatically.

Otherwise, use AskUserQuestion:

  • Question: "Run deep engineering review? Full interactive review of architecture, code quality, tests, and performance — goes beyond Phase 2.5."
  • Options: "Continue (recommended)" / "Skip"

If continue: Run /plan-eng-review via Skill tool with $PLAN_PATH as arguments.

If skip: Proceed.

Phase 5: Design Review (opt-out)

If --skip-reviews: skip to Phase 7.

If --yes: run automatically.

Otherwise, use AskUserQuestion:

  • Question: "Run deep design review? Scores 7 UI/UX dimensions 0-10 and fixes the plan."
  • Options: "Continue (recommended)" / "Skip"

If continue: Run /plan-design-review via Skill tool with $PLAN_PATH as arguments.

If skip: Proceed.

Phase 6: Multi-Review on Plan (opt-out)

If --skip-reviews: skip to Phase 7.

If --yes: run automatically.

Otherwise, use AskUserQuestion:

  • Question: "Run specialized reviewers (security, performance, architecture) on the plan? Catches issues before code is written."
  • Options: "Continue (recommended)" / "Skip"

If continue: Run /multi-review via Skill tool with --plan $PLAN_PATH as arguments.

If skip: Proceed.

Phase 7: Decompose (opt-out)

If --no-decompose: skip to Pipeline Summary.

Check if Linear MCP is available (try calling list_teams). If unavailable, skip silently. Log: "Linear MCP not available — skipping decomposition."

Check if /spec already decomposed during Phase 3 (look for Linear issues associated with the plan). If already decomposed, skip.

If --yes: run decomposition automatically.

Otherwise, use AskUserQuestion:

  • Question: "Decompose the plan into Linear issues for tracking and dispatch?"
  • Options: "Yes, create issues (recommended)" / "No, plan file is sufficient"

If yes: Run /spec via Skill tool with $PLAN_PATH as arguments. Spec will detect the existing plan and offer decomposition.

If no: Proceed.

Pipeline Summary

Output a completion summary:

═══════════════════════════════════════════
DEEP-PLAN COMPLETE
═══════════════════════════════════════════
Feature: $FEATURE
Plan: $PLAN_PATH

PHASES COMPLETED:
  1. Product Review     ✅
  2. Design Shotgun     ✅ / ⏭ skipped
  3. Spec               ✅  (Phase 2.5 inline review ran)
  4. Eng Review         ✅ / ⏭ skipped
  5. Design Review      ✅ / ⏭ skipped
  6. Multi-Review       ✅ / ⏭ skipped
  7. Decomposition      ✅ / ⏭ skipped / ⚠ Linear unavailable

NEXT STEPS:
  /orient → /dispatch         (parallel execution)
  /start-task <id>            (solo execution)
  /auto-run --through <id>    (autonomous execution)
═══════════════════════════════════════════

Rules

  • NEVER duplicate sub-skill content. This skill sequences, it does not replace.
  • ALWAYS pass $FEATURE or $PLAN_PATH to sub-skills so they have context.
  • NEVER skip Phase 1 (product review) or Phase 3 (spec). They are the backbone. If you don't want product review, use /spec directly.
  • Each sub-skill handles its own internal interaction (questions, modes, outputs). This skill only manages transitions between them.
  • If a sub-skill fails or the user exits it early, stop the pipeline. Do not silently continue.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most plan spec skills give in ~1.5k tokens

Counted across 1,360 of the 2,617 authors here whose files we hold, read 2026-09-06

  • Ask one question at a timein 73 of 1360
  • Write the spec using the templatein 22 of 1360
  • Ask clarifying questions if neededin 19 of 1360, across 18 files
  • Wait for user confirmation before proceedingin 19 of 1360
  • Save plans to the plans directoryin 17 of 1360, across 13 files
  • Check for product marketing context firstin 16 of 1360, across 5 files
  • Read the plan file completelyin 16 of 1360
  • Order tasks by dependencyin 16 of 1360
  • Gather context from the conversationin 15 of 1360, across 9 files
  • Explore the codebase instead of askingin 15 of 1360, across 13 files
  • Wait for explicit user approvalin 14 of 1360, across 13 files
  • Quiz the user on the breakdownin 13 of 1360, across 7 files

Said here and by no other author read

  • Run product review as the first phase
  • Pass feature or plan path to sub-skills
  • Stop pipeline if a sub-skill fails
  • Check if Linear MCP is available before decomposing

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.