speckit-plan

name	speckit-plan
description	Execute the implementation planning workflow using the plan template to generate design artifacts.
compatibility	Requires spec-kit project structure with .specify/ directory
metadata	{"author":"github-spec-kit","source":"preset:companion-standard"}
user-invocable	true
disable-model-invocation	false

Speckit Plan Skill

User Input

$ARGUMENTS

You MUST consider the user input before proceeding (if not empty).

Pre-Execution Checks

Check for extension hooks (before planning):

• Check if .specify/extensions.yml exists in the project root.
• If it exists, read it and look for entries under the hooks.before_plan key
• If the YAML cannot be parsed or is invalid, skip hook checking silently and continue normally
• Filter out hooks where enabled is explicitly false. Treat hooks without an enabled field as enabled by default.
• For each remaining hook, do not attempt to interpret or evaluate hook condition expressions:
  • If the hook has no condition field, or it is null/empty, treat the hook as executable
  • If the hook defines a non-empty condition, skip the hook and leave condition evaluation to the HookExecutor implementation
• When constructing slash commands from hook command names, replace dots (.) with hyphens (-). For example, speckit.git.commit → /speckit-git-commit.
• For each executable hook, output the following based on its optional flag:
  • Optional hook (optional: true):

    ## Extension Hooks
    
    **Optional Pre-Hook**: {extension}
    Command: `/{command}`
    Description: {description}
    
    Prompt: {prompt}
    To execute: `/{command}`
    

  • Mandatory hook (optional: false):

    ## Extension Hooks
    
    **Automatic Pre-Hook**: {extension}
    Executing: `/{command}`
    EXECUTE_COMMAND: {command}
    
    Wait for the result of the hook command before proceeding to the Outline.
    

• If no hooks are registered or .specify/extensions.yml does not exist, skip silently

Outline

1. Setup: Run .specify/scripts/bash/setup-plan.sh --json from repo root and parse JSON for FEATURE_SPEC, IMPL_PLAN, SPECS_DIR, BRANCH. For single quotes in args like "I'm Groot", use escape syntax: e.g 'I'''m Groot' (or double-quote if possible: "I'm Groot").

2. Load context: Read FEATURE_SPEC and .specify/memory/constitution.md. Load IMPL_PLAN template (already copied).

3. Execute plan workflow: Follow the structure in IMPL_PLAN template to:

   • Fill Technical Context (mark unknowns as "NEEDS CLARIFICATION")
   • Fill Constitution Check section from constitution
   • Evaluate gates (ERROR if violations unjustified)
   • Phase 0: Generate research.md (resolve all NEEDS CLARIFICATION)
   • Phase 1: Generate data-model.md, contracts/, quickstart.md
   • Phase 1: Update agent context by running the agent script
   • Re-evaluate Constitution Check post-design

4. Stop and report: Command ends after Phase 2 planning. Report branch, IMPL_PLAN path, and generated artifacts.

5. Check for extension hooks: After reporting, check if .specify/extensions.yml exists in the project root.

   • If it exists, read it and look for entries under the hooks.after_plan key
   • If the YAML cannot be parsed or is invalid, skip hook checking silently and continue normally
   • Filter out hooks where enabled is explicitly false. Treat hooks without an enabled field as enabled by default.
   • For each remaining hook, do not attempt to interpret or evaluate hook condition expressions:
     • If the hook has no condition field, or it is null/empty, treat the hook as executable
     • If the hook defines a non-empty condition, skip the hook and leave condition evaluation to the HookExecutor implementation
   • When constructing slash commands from hook command names, replace dots (.) with hyphens (-). For example, speckit.git.commit → /speckit-git-commit.
   • For each executable hook, output the following based on its optional flag:
     • Optional hook (optional: true):

       ## Extension Hooks
       
       **Optional Hook**: {extension}
       Command: `/{command}`
       Description: {description}
       
       Prompt: {prompt}
       To execute: `/{command}`
       

     • Mandatory hook (optional: false):

       ## Extension Hooks
       
       **Automatic Hook**: {extension}
       Executing: `/{command}`
       EXECUTE_COMMAND: {command}
       

   • If no hooks are registered or .specify/extensions.yml does not exist, skip silently

Phases

Phase 0: Outline & Research

1. Extract unknowns from Technical Context above:

   • For each NEEDS CLARIFICATION → research task
   • For each dependency → best practices task
   • For each integration → patterns task

2. Generate and dispatch research agents:

   For each unknown in Technical Context:
     Task: "Research {unknown} for {feature context}"
   For each technology choice:
     Task: "Find best practices for {tech} in {domain}"
   

3. Consolidate findings in research.md using format:

   • Decision: [what was chosen]
   • Rationale: [why chosen]
   • Alternatives considered: [what else evaluated]

Output: research.md with all NEEDS CLARIFICATION resolved

Phase 1: Design & Contracts

Prerequisites: research.md complete

1. Extract entities from feature spec → data-model.md:

   • Entity name, fields, relationships
   • Validation rules from requirements
   • State transitions if applicable

2. Define interface contracts (if project has external interfaces) → /contracts/:

   • Identify what interfaces the project exposes to users or other systems
   • Document the contract format appropriate for the project type
   • Examples: public APIs for libraries, command schemas for CLI tools, endpoints for web services, grammars for parsers, UI contracts for applications
   • Skip if project is purely internal (build scripts, one-off tools, etc.)

3. Agent context update:

   • Update the plan reference between the <!-- SPECKIT START --> and <!-- SPECKIT END --> markers in CLAUDE.md to point to the plan file created in step 1 (the IMPL_PLAN path)

Output: data-model.md, /contracts/*, quickstart.md, updated agent context file

Key rules

• Use absolute paths for filesystem operations; use project-relative paths for references in documentation and agent context files
• ERROR on gate failures or unresolved clarifications

Timing — keep .spec-context.json honest

These rules apply to every Companion profile command. The extension records lifecycle timing with its own scripts wherever it can; these rules keep anything you append consistent with that and accurate for any dispatcher (terminal, IDE chat, or the GUI). The model is finish-only: each task and each substep records a single finish event, and its duration is the gap to the previous finish (or the step's start). Never a start+complete pair for a task or substep — a pair stamped at one instant is what produces 0s ticks and bursts.

• Never hand-edit .spec-context.json. Record every finish by running the writer script, never by editing the JSON file yourself — a hand-authored edit is what corrupts the file (a duplicated status key). The script stamps the real clock, writes atomically, and is idempotent. The commands below are the only way you touch timing.

• Self-close — but not specify or implement. When your own work for plan, tasks, clarify, or analyze ends, record the step finish (feature dir from .specify/feature.json):

  python3 .specify/extensions/companion/scripts/write-context.py --feature-dir <feature_dir> --step <this step> --finish --by ai
  

  --finish appends a single step-level complete and touches nothing else (it leaves status/currentStep to the lifecycle hooks). Do NOT self-close specify or implement: the extension closes those itself (specify from its own command, implement from the end-of-step hook), so an ai complete there would duplicate it.

• Substeps — one finish each, via the script. For each substep boundary (plan: research, design; tasks: generate), the moment that substep ends, run:

  python3 .specify/extensions/companion/scripts/write-context.py --feature-dir <feature_dir> --step <step> --substep <name> --finish --by ai
  

  One call per substep, each stamped with its own real clock at the moment it finishes — never two substeps in one batch, never a separate start. The delta between consecutive finishes is each substep's duration.

• Implement — finishing a task is logging it (finish-only). Recording a task's finish is the closing action of that task, done the instant its work is complete and before you start the next one — not a bookkeeping pass you batch at the end of a phase. The closing action is a single append (feature dir from .specify/feature.json):

  python3 .specify/extensions/companion/scripts/write-context.py --feature-dir <feature_dir> --task <TaskID> --kind complete --by ai --did "<one-line summary of what this task did>" --files "<comma,separated,files,touched>" --append
  

  --append writes one line to .spec-context.events.jsonl and does not read or rewrite the shared .spec-context.json, so it never hits the "read the file first" retry and parallel workers can each append their own finish at the same time without contending — the line carries its own real timestamp (date -u is stamped by the script). The --did/--files flags ride along so the Activity panel's Tasks card is populated from the script. Do NOT hand-edit the - [ ] checkbox in tasks.md — the script owns it: materialize flips it to - [x] from your appended finish, so a fanned-out subagent only appends and never touches the shared tasks.md. Do NOT hand-author per-task JSON and do NOT write a per-task start.

  Then fold the appended lines into .spec-context.json — run this once per wave (after the wave reconciles) and again when the step ends:

  python3 .specify/extensions/companion/scripts/write-context.py --feature-dir <feature_dir> --materialize
  

  --materialize is the one read-modify-write: it folds the finishes into the panel and checks off the matching tasks.md boxes for every journaled task, idempotently (re-folding never double-counts), so running it per wave keeps the GUI current without re-serializing the work. The end-of-step hook is a backstop that materializes anything you didn't fold and fills any task you didn't journal. What's trustworthy here is the per-task summary (did/files) and the order tasks completed, plus the step-level start→complete span, which the scripts stamp exactly. The per-task timestamps are best-effort — a single agent logs a task right after building it, so they reflect when you recorded it, not a precisely measured duration; that's fine, the summaries are the point. Still, record each finish as you go, wave by wave rather than dumping every task at the very end — a per-wave cadence keeps the panel live and the ordering true.

• Never write the next step's start. Only the next command appends the next step's start entry; writing it here makes the viewer render a phantom "Generating …".