Agent skill

Workflow Plan

by catlog22 in catlog22/Claude-Code-Workflow

Planning pipeline with multi-mode routing (plan/verify/replan).

MITAuto-check: notesAgent Workflows

Install Workflow Plan

skills CLI
$ npx skills add catlog22/Claude-Code-Workflow --skill workflow-plan -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install catlog22/Claude-Code-Workflow workflow-plan --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/catlog22/Claude-Code-Workflow.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.codex/skills/workflow-plan .claude/skills/workflow-plan && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
workflow-plan
GitHub stars
2.1k
Token cost
~5.5k tokens
SKILL.md length
249 words
Files
1
Skills in repo
82
Repo updated
First seen
Licence
MIT

At a glance

Planning pipeline with multi-mode routing (plan/verify/replan).

  • Works in 6 steps: Session Discovery (Plan Mode) → Context Gathering (spawn_agent) → Conflict Resolution (Conditional) → …
  • Tasks that involve Subagents
  • SKILL.md covers Auto Mode, Usage, Overview and Data Flow, plus 4 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Workflow Plan is an agent skill from catlog22/Claude-Code-Workflow. Planning pipeline with multi-mode routing (plan/verify/replan). Session discovery → context gathering (spawnagent) → conditional conflict resolution → task generation (spawnagent or N+1 parallel agents) → plan verification → interactive replan. Produces IMPLPLAN.md, task JSONs, TODOLIST.md.

Its SKILL.md is about 5.5k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Agent Workflows, covering Subagents, Task management and Planning. The repository describes itself as: JSON-driven multi-agent cadence-team development framework with intelligent CLI orchestration (Gemini/Qwen/Codex), context-first architecture, and automated workflow execution. The licence is MIT.

When your agent uses it

  • Tasks that involve Subagents
  • Tasks that involve Task management
  • Tasks that involve Planning

Example prompts

  • “/workflow-plan”

Requirements

  • Pre-approved tools (allowed-tools): spawn_agent, wait_agent, send_message, followup_task, close_agent, request_user_input, Read, Write, Edit, Bash, Glob, Grep

Workflow steps

6 steps, taken from the step headings in SKILL.md.

  1. Session Discovery (Plan Mode)
  2. Context Gathering (spawn_agent)
  3. Conflict Resolution (Conditional)
  4. Task Generation (spawn_agent)
  5. Plan Verification (Verify Mode)
  6. Interactive Replan (Replan Mode)

What it can do on your machine

Read from SKILL.md and the folder at commit 07491b0. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • spawn_agent
    • wait_agent
    • send_message
    • followup_task
    • close_agent
    • request_user_input
    • Read
    • Write
    • Edit
    • Bash

    …and 2 more on the same allowed-tools line.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are javascript and bash).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Workflow Plan loads about 5.5k tokens when it runs. Until then it costs about 77 tokens; SKILL.md has 249 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~77
When it runs · the whole SKILL.md, loaded when a task matches
~5.5k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

  • NotePre-approves every shell command (allowed-tools: Bash)SKILL.md
    allowed-tools: spawn_agent, wait_agent, send_message, followup_task, close_agent, request_user_input, Read, Write, 

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from catlog22/Claude-Code-Workflow at commit 07491b0, republished under its MIT licence (© catlog22). 249 words, ~5,503 tokens.

Download SKILL.mdSave it as .claude/skills/workflow-plan/SKILL.md (or your agent's skills folder).
name
workflow-plan
description
Planning pipeline with multi-mode routing (plan/verify/replan). Session discovery → context gathering (spawn_agent) → conditional conflict resolution → task generation (spawn_agent or N+1 parallel agents) → plan verification → interactive replan. Produces IMPL_PLAN.md, task JSONs, TODO_LIST.md.
allowed-tools
spawn_agent, wait_agent, send_message, followup_task, close_agent, request_user_input, Read, Write, Edit, Bash, Glob, Grep
argument-hint
[-y|--yes] [--session ID] "task description" | verify [--session ID] | replan [--session ID] [IMPL-N] "changes"

Auto Mode

When --yes or -y: Skip all confirmations, use defaults, auto-verify. This skill is planning-only — it NEVER executes implementation. Output is the plan for user review.

Workflow Plan

Usage

bash
# Plan mode (default)
$workflow-plan "Build authentication system with JWT and OAuth"
$workflow-plan -y "Add rate limiting to API endpoints"
$workflow-plan --session WFS-auth "Extend with 2FA support"

# Verify mode
$workflow-plan verify --session WFS-auth
$workflow-plan verify

# Replan mode
$workflow-plan replan --session WFS-auth "Change from JWT to session-based auth"
$workflow-plan replan --session WFS-auth IMPL-3 "Split into two smaller tasks"

Flags:

  • -y, --yes: Skip all confirmations (auto mode)
  • --session ID: Use specific session

Overview

Multi-mode planning pipeline using subagent coordination. Plan mode runs 4 sequential phases with conditional branching; verify and replan modes operate on existing plans.

┌──────────────────────────────────────────────────────────────────┐
│                    WORKFLOW PLAN PIPELINE                         │
├──────────────────────────────────────────────────────────────────┤
│                                                                    │
│  Mode Detection: plan | verify | replan                           │
│                                                                    │
│  ═══ Plan Mode (default) ═══                                      │
│                                                                    │
│  Phase 1: Session Discovery                                       │
│     ├─ Create or find workflow session                            │
│     └─ Initialize planning-notes.md                               │
│                                                                    │
│  Phase 2: Context Gathering (spawn_agent: context-search-agent)  │
│     ├─ Codebase analysis → context-package.json                  │
│     └─ Conflict risk assessment                                   │
│                                                                    │
│  Phase 3: Conflict Resolution (conditional: risk ≥ medium)       │
│     ├─ CLI-driven conflict analysis                               │
│     └─ User-selected resolution strategies                        │
│                                                                    │
│  Phase 4: Task Generation (spawn_agent: action-planning-agent)   │
│     ├─ Single module → 1 agent                                    │
│     ├─ Multi-module → N+1 parallel agents                        │
│     └─ Output: IMPL_PLAN.md + task JSONs + TODO_LIST.md          │
│                                                                    │
│  Plan Confirmation Gate (PLANNING ENDS HERE)                     │
│     ├─ "Verify Plan" → Phase 5                                   │
│     ├─ "Done" → Display next-step command for user               │
│     └─ "Review Status" → Display inline                          │
│                                                                    │
│  ═══ Verify Mode ═══                                              │
│  Phase 5: Plan Verification (spawn_agent: cli-explore-agent)     │
│     └─ 10-dimension analysis → PLAN_VERIFICATION.md              │
│                                                                    │
│  ═══ Replan Mode ═══                                              │
│  Phase 6: Interactive Replan                                      │
│     └─ Clarification → Impact → Backup → Apply → Verify          │
│                                                                    │
└──────────────────────────────────────────────────────────────────┘

Data Flow

User Input (task description)
    │
    ↓ [Convert to GOAL/SCOPE/CONTEXT]
    │
Phase 1 ──→ sessionId, planning-notes.md
    │
Phase 2 ──→ context-package.json, conflictRisk
    │
    ├── conflictRisk ≥ medium ──→ Phase 3 ──→ modified artifacts
    └── conflictRisk < medium ──→ skip Phase 3
    │
Phase 4 ──→ IMPL_PLAN.md, plan.json, task JSONs, TODO_LIST.md
    │
    ├── Verify → Phase 5 → PLAN_VERIFICATION.md
    ├── Execute → workflow-execute skill
    └── Review → inline display

Session Structure

.workflow/active/WFS-{session}/
├── workflow-session.json              # Session metadata
├── planning-notes.md                  # Accumulated context across phases
├── IMPL_PLAN.md                       # Implementation plan (human-readable)
├── plan.json                          # Structured plan overview (machine-readable)
├── TODO_LIST.md                       # Task checklist
├── .task/                             # Task definitions
│   ├── IMPL-1.json
│   └── IMPL-N.json
└── .process/
    ├── context-package.json           # Phase 2 output
    ├── conflict-resolution.json       # Phase 3 output (conditional)
    ├── PLAN_VERIFICATION.md           # Phase 5 output
    └── backup/                        # Phase 6 backups

Implementation

Session Initialization
javascript
const getUtc8ISOString = () => new Date(Date.now() + 8 * 60 * 60 * 1000).toISOString()

// Parse flags
const AUTO_YES = $ARGUMENTS.includes('--yes') || $ARGUMENTS.includes('-y')
const sessionMatch = $ARGUMENTS.match(/--session\s+(\S+)/)
const existingSessionId = sessionMatch ? sessionMatch[1] : null

// Mode detection
const cleanArgs = $ARGUMENTS
  .replace(/--yes|-y|--session\s+\S+/g, '').trim()

let mode = 'plan'
if (cleanArgs.startsWith('verify')) mode = 'verify'
else if (cleanArgs.startsWith('replan')) mode = 'replan'

const taskDescription = cleanArgs
  .replace(/^(verify|replan)\s*/, '')
  .replace(/^["']|["']$/g, '')
  .trim()

// Extract replan task ID if present
const replanTaskMatch = taskDescription.match(/^(IMPL-\d+(?:\.\d+)?)\s+(.+)/)
const replanTaskId = replanTaskMatch ? replanTaskMatch[1] : null
const replanDescription = replanTaskMatch ? replanTaskMatch[2] : taskDescription

Phase 1: Session Discovery (Plan Mode)

Objective: Create or find workflow session, initialize planning notes.

javascript
if (mode !== 'plan') {
  // verify/replan: locate existing session
  // → Jump to Phase 5 or Phase 6
}

let sessionId, sessionFolder

if (existingSessionId) {
  sessionId = existingSessionId
  sessionFolder = `.workflow/active/${sessionId}`
  if (!Bash(`test -d "${sessionFolder}" && echo yes`).trim()) {
    console.log(`ERROR: Session ${sessionId} not found`)
    return
  }
} else {
  // Auto-detect from .workflow/active/ or create new
  const sessions = Bash(`ls -d .workflow/active/WFS-* 2>/dev/null`).trim().split('\n').filter(Boolean)

  if (sessions.length === 0 || taskDescription) {
    // Create new session
    const slug = taskDescription.toLowerCase()
      .replace(/[^a-z0-9\u4e00-\u9fa5]+/g, '-').substring(0, 40)
    sessionId = `WFS-${slug}`
    sessionFolder = `.workflow/active/${sessionId}`
    Bash(`mkdir -p "${sessionFolder}/.task" "${sessionFolder}/.process" "${sessionFolder}/.summaries"`)

    Write(`${sessionFolder}/workflow-session.json`, JSON.stringify({
      session_id: sessionId,
      status: 'planning',
      created_at: getUtc8ISOString(),
      task_description: taskDescription
    }, null, 2))
  } else if (sessions.length === 1) {
    sessionId = sessions[0].split('/').pop()
    sessionFolder = sessions[0]
  } else {
    // Multiple sessions — ask user
    if (AUTO_YES) {
      sessionFolder = sessions[0]
      sessionId = sessions[0].split('/').pop()
    } else {
      const answer = functions.request_user_input({
        questions: [{
          question: "Multiple sessions found. Select one:",
          header: "Session",
          options: sessions.slice(0, 4).map(s => ({
            label: s.split('/').pop(),
            description: s
          }))
        }]
      })
      sessionId = answer.Session
      sessionFolder = `.workflow/active/${sessionId}`
    }
  }
}

// Initialize planning-notes.md
const structuredDesc = `GOAL: ${taskDescription}\nSCOPE: Core implementation\nCONTEXT: New development`
Write(`${sessionFolder}/planning-notes.md`, `# Planning Notes\n\n## User Intent\n${structuredDesc}\n`)

console.log(`Session: ${sessionId}`)

Phase 2: Context Gathering (spawn_agent)

Objective: Gather project context, assess conflict risk.

javascript
console.log(`\n## Phase 2: Context Gathering\n`)

const ctxAgent = spawn_agent({
  agent_type: "context_search_agent",
  instruction: `
Gather implementation context for planning.

**Session**: ${sessionFolder}
**Task**: ${taskDescription}
**Mode**: PLAN

### Steps
1. Analyze project structure (package.json, tsconfig, etc.)
2. Search for existing similar implementations
3. Identify integration points and dependencies
4. Assess conflict risk with existing code
5. Generate context package

### Output
Write context package to: ${sessionFolder}/.process/context-package.json
Format: {
  "critical_files": [...],
  "patterns": [...],
  "dependencies": [...],
  "integration_points": [...],
  "conflict_risk": "none" | "low" | "medium" | "high",
  "conflict_areas": [...],
  "constraints": [...]
}
`
})

wait_agent({ timeout_ms: 1800000 })  // 30 minutes
close_agent({ target: ctxAgent })

// Parse outputs
const contextPkg = JSON.parse(Read(`${sessionFolder}/.process/context-package.json`) || '{}')
const conflictRisk = contextPkg.conflict_risk || 'none'
const contextPath = `${sessionFolder}/.process/context-package.json`

// Update planning-notes.md
Edit(`${sessionFolder}/planning-notes.md`, {
  oldText: '## User Intent',
  newText: `## Context Findings
- Critical files: ${(contextPkg.critical_files || []).join(', ')}
- Conflict risk: ${conflictRisk}
- Constraints: ${(contextPkg.constraints || []).join('; ')}

## User Intent`
})

console.log(`  Context gathered. Conflict risk: ${conflictRisk}`)

Phase 3: Conflict Resolution (Conditional)

Objective: Detect and resolve conflicts when risk ≥ medium.

javascript
if (['medium', 'high'].includes(conflictRisk)) {
  console.log(`\n## Phase 3: Conflict Resolution (risk: ${conflictRisk})\n`)

  Bash({
    command: `ccw cli -p "PURPOSE: Analyze and resolve conflicts between planned changes and existing codebase.
TASK:
  • Read context package for conflict areas
  • Analyze each conflict area in detail
  • Propose resolution strategies (refactor, adapt, isolate, defer)
  • For each conflict: assess impact and recommend approach
MODE: analysis
CONTEXT: @**/*
EXPECTED: JSON: {conflicts: [{area, severity, description, strategy, impact}], summary: string}
CONSTRAINTS: Focus on ${(contextPkg.conflict_areas || []).join(', ')}

TASK DESCRIPTION: ${taskDescription}" --tool gemini --mode analysis --rule analysis-diagnose-bug-root-cause`,
    run_in_background: true
  })
  // Wait for CLI → conflicts[]

  if (!AUTO_YES && conflicts.length > 0) {
    // Present conflicts and let user select strategies
    console.log(`\n### Conflicts Found: ${conflicts.length}\n`)
    conflicts.forEach((c, i) => {
      console.log(`${i + 1}. [${c.severity}] ${c.area}: ${c.description}`)
      console.log(`   Strategy: ${c.strategy} | Impact: ${c.impact}`)
    })

    const answer = functions.request_user_input({
      questions: [{
        question: "Accept conflict resolution strategies?",
        header: "Conflicts",
        options: [
          { label: "Accept All", description: "Apply all recommended strategies" },
          { label: "Review Each", description: "Approve strategies individually" },
          { label: "Skip", description: "Proceed without resolving" }
        ]
      }]
    })
  }

  // Write resolution
  Write(`${sessionFolder}/.process/conflict-resolution.json`,
    JSON.stringify({ conflicts, resolved_at: getUtc8ISOString() }, null, 2))

  // Update planning-notes
  // Append conflict decisions to planning-notes.md
} else {
  console.log(`  Conflict risk: ${conflictRisk} — skipping Phase 3`)
}

Phase 4: Task Generation (spawn_agent)

Objective: Generate IMPL_PLAN.md, task JSONs, TODO_LIST.md.

Steps:

  1. Determine Planning Strategy

    javascript
    console.log(`\n## Phase 4: Task Generation\n`)
    
    // Detect module count from context
    const modules = contextPkg.integration_points?.map(p => p.module).filter(Boolean) || []
    const uniqueModules = [...new Set(modules)]
    const isMultiModule = uniqueModules.length >= 2
  2. Single Module → One Agent

    javascript
    if (!isMultiModule) {
      const planAgent = spawn_agent({
        agent_type: "action_planning_agent",
        instruction: `

Generate implementation plan and task JSONs.

Session: ${sessionFolder} Task: ${taskDescription} Context: ${contextPath} Planning Notes: ${sessionFolder}/planning-notes.md ${contextPkg.conflict_risk === 'medium' || contextPkg.conflict_risk === 'high' ? **Conflict Resolution**: ${sessionFolder}/.process/conflict-resolution.json : ''}

Output Requirements
  1. IMPL_PLAN.md at ${sessionFolder}/IMPL_PLAN.md

    • Section 1: Requirements Summary
    • Section 2: Architecture Decisions
    • Section 3: Task Breakdown (with dependencies)
    • Section 4: Implementation Strategy (Sequential/Parallel/Phased)
    • Section 5: Risk Assessment
  2. plan.json at ${sessionFolder}/plan.json

    • {task_ids[], recommended_execution, complexity, shared_context}
  3. Task JSONs at ${sessionFolder}/.task/IMPL-{N}.json

    • {id, title, description, depends_on[], convergence, meta: {type, agent}}
  4. TODO_LIST.md at ${sessionFolder}/TODO_LIST.md

    • Checkbox format: - [ ] IMPL-{N}: {title} ` })

      wait_agent({ timeout_ms: 1800000 }) // 30 minutes close_agent({ target: planAgent }) }

  5. Multi-Module → N+1 Parallel Agents

    javascript
    if (isMultiModule) {
      const prefixes = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
      const moduleAgents = []
    
      // Spawn N module planners in parallel
      for (let i = 0; i < uniqueModules.length; i++) {
        const prefix = prefixes[i]
        const mod = uniqueModules[i]
    
        const agentId = spawn_agent({
          agent_type: "action_planning_agent",
          instruction: `

Plan module: ${mod} (prefix: ${prefix})

Session: ${sessionFolder} Module: ${mod} Context: ${contextPath} Task ID prefix: ${prefix} (e.g., ${prefix}1, ${prefix}2, ...)

Generate task JSONs for this module only. Output to: ${sessionFolder}/.task/${prefix}{N}.json Mark cross-module dependencies as CROSS::${'{module}'}::${'{task}'} ` }) moduleAgents.push({ id: agentId, module: mod, prefix }) }

 // Wait for all module planners
 wait_agent({ timeout_ms: 1800000 })  // 30 minutes
 moduleAgents.forEach(a => close_agent({ target: a.id }))

 // +1 Coordinator: integrate all modules
 const coordAgent = spawn_agent({
   agent_type: "action_planning_agent",
   instruction: `

Integrate ${uniqueModules.length} module plans into unified IMPL_PLAN.md.

Session: ${sessionFolder} Modules: ${uniqueModules.join(', ')} Task Directory: ${sessionFolder}/.task/

Show full SKILL.md (268 more words)Show less
Steps
  1. Read all module task JSONs from .task/

  2. Resolve CROSS:: dependencies (replace with actual task IDs)

  3. Generate unified IMPL_PLAN.md, plan.json, TODO_LIST.md

  4. Renumber task IDs to sequential IMPL-1, IMPL-2, ... ` })

    wait_agent({ timeout_ms: 1800000 }) // 30 minutes close_agent({ target: coordAgent }) }

  5. Plan Confirmation Gate

    javascript
    // Validate outputs exist
    const planExists = Bash(`test -f "${sessionFolder}/IMPL_PLAN.md" && echo yes`).trim() === 'yes'
    const taskCount = parseInt(Bash(`ls ${sessionFolder}/.task/IMPL-*.json 2>/dev/null | wc -l`).trim()) || 0
    
    console.log(`\n## Plan Generated\n`)
    console.log(`  Tasks: ${taskCount}`)
    console.log(`  Plan: ${sessionFolder}/IMPL_PLAN.md`)
    
    if (AUTO_YES) {
      // Auto-verify plan quality
      console.log(`  [--yes] Auto-verifying plan...`)
      // → Fall through to Phase 5
    } else {
      const nextStep = functions.request_user_input({
        questions: [{
          question: "Plan generated. What's next?",
          header: "Next Step",
          options: [
            { label: "Verify Plan (Recommended)", description: "Run quality verification" },
            { label: "Done", description: "Planning complete — show next-step command" },
            { label: "Review Status", description: "Display plan summary inline" }
          ]
        }]
      })
    
      if (nextStep['Next Step'] === 'Done') {
        console.log(`\nPlanning complete. To execute, run: $workflow-execute --session ${sessionId}`)
        return  // STOP — this skill is planning-only
      }
      if (nextStep['Next Step'] === 'Review Status') {
        const plan = Read(`${sessionFolder}/IMPL_PLAN.md`)
        console.log(plan)
        return  // STOP — this skill is planning-only
      }
      // Verify → continue to Phase 5
    }

Phase 5: Plan Verification (Verify Mode)

Objective: Read-only multi-dimensional plan analysis.

javascript
if (mode === 'verify' || /* auto-verify from Phase 4 */) {
  console.log(`\n## Phase 5: Plan Verification\n`)

  // Find session if in verify mode entry
  if (mode === 'verify' && !sessionFolder) {
    // Session discovery (same logic as Phase 1)
  }

  Bash({
    command: `ccw cli -p "PURPOSE: Verify implementation plan quality across 10 dimensions. Read-only analysis. Success = actionable quality gate recommendation.
TASK:
  • A: User Intent Alignment — does plan match original goal?
  • B: Requirements Coverage — are all requirements addressed?
  • C: Consistency Validation — internal consistency of plan
  • D: Dependency Integrity — valid dependency chain
  • E: Synthesis Alignment — matches brainstorm artifacts (if exist)
  • F: Task Specification Quality — clear, actionable, testable
  • G: Duplication Detection — no overlapping tasks
  • H: Feasibility Assessment — realistic scope and effort
  • I: Constraints Compliance — respects stated constraints
  • J: Context Validation — planning-notes consistent with plan
MODE: analysis
CONTEXT: @${sessionFolder}/IMPL_PLAN.md @${sessionFolder}/.task/**/*.json @${sessionFolder}/planning-notes.md @${sessionFolder}/TODO_LIST.md
EXPECTED: Structured report with: per-dimension score (PASS/WARN/FAIL), issues list, quality gate (BLOCK_EXECUTION/PROCEED_WITH_FIXES/PROCEED_WITH_CAUTION/PROCEED)
CONSTRAINTS: Read-only | No file modifications | Be specific about issues" --tool gemini --mode analysis --rule analysis-review-architecture --cd "${sessionFolder}"`,
    run_in_background: true
  })
  // Wait for CLI → verification report

  Write(`${sessionFolder}/.process/PLAN_VERIFICATION.md`, verificationReport)

  console.log(`  Quality gate: ${qualityGate}`)
  console.log(`  Report: ${sessionFolder}/.process/PLAN_VERIFICATION.md`)

  console.log(`\nPlanning complete. To execute, run: $workflow-execute --session ${sessionId}`)
  // STOP — this skill is planning-only, NEVER proceed to execution
}

Phase 6: Interactive Replan (Replan Mode)

Objective: Modify existing plan based on new requirements.

javascript
if (mode === 'replan') {
  console.log(`\n## Phase 6: Interactive Replan\n`)

  // Find session
  if (!sessionFolder) {
    // Session discovery logic
  }

  const scope = replanTaskId ? 'task' : 'session'
  console.log(`  Scope: ${scope}${replanTaskId ? ` (${replanTaskId})` : ''}`)
  console.log(`  Changes: ${replanDescription}`)

  // 1. Backup current plan
  Bash(`mkdir -p "${sessionFolder}/.process/backup" && cp "${sessionFolder}/IMPL_PLAN.md" "${sessionFolder}/.process/backup/IMPL_PLAN-$(date +%Y%m%d%H%M%S).md"`)

  // 2. Replan via agent
  const replanAgent = spawn_agent({
    agent_type: "action_planning_agent",
    instruction: `
Replan ${scope === 'task' ? `task ${replanTaskId}` : 'entire session'}.

**Session**: ${sessionFolder}
**Current Plan**: ${sessionFolder}/IMPL_PLAN.md
**Current Tasks**: ${sessionFolder}/.task/
**Changes Requested**: ${replanDescription}
${replanTaskId ? `**Target Task**: ${sessionFolder}/.task/${replanTaskId}.json` : ''}

### Steps
1. Read current plan and task JSONs
2. Analyze impact of requested changes
3. Modify affected tasks (update/add/remove)
4. Update IMPL_PLAN.md with change annotations
5. Update TODO_LIST.md
6. Generate change summary

### Constraints
- Preserve completed tasks
- Minimize changes to unaffected tasks
- Maintain dependency integrity
`
  })

  wait_agent({ timeout_ms: 1800000 })  // 30 minutes
  close_agent({ target: replanAgent })

  console.log(`  Replan complete. Review: ${sessionFolder}/IMPL_PLAN.md`)
}

Error Handling

ErrorRecovery
No active sessionGuide: run $workflow-plan "description" first
Context gathering timeoutRetry with reduced scope
Conflict analysis failureSkip Phase 3, proceed with warning
Task generation failureRetry agent with simplified context
Plan verification failureDisplay partial results
Session not found (verify/replan)List available sessions

Compact Protection

COMPACT DIRECTIVE: If context compression occurs:

  • Re-read planning-notes.md for accumulated context
  • Re-read workflow-session.json for session state
  • Phase 4 is the most token-intensive — if compressed, re-read task JSONs on demand

© catlog22, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .codex/skills/workflow-plan of catlog22/Claude-Code-Workflow.

Open the folder on GitHubat commit 07491b0

Compare with similar skills

Workflow Plan next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Workflow Plan compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Workflow Plan this skillcatlog22/Claude-Code-Workflow2.1k—~5.5kAutomated safety check: NotesMIT
Subagent Driven DevelopmentAsvarox/allkaraoke26137 repos~1.2kAutomated safety check: PassNone
Executing PlansGanyuanRan/Aegis1.3k1 repos~2.3kAutomated safety check: PassMIT
Execumputun/cc-thingz483—~8kAutomated safety check: PassMIT
Autopilot End-to-End Buildernick-vels/skills403—~2.5kAutomated safety check: NotesMIT
Workflow Orchestrationvxcozy/workflow-orchestration115—~1kAutomated safety check: PassMIT

Similar skills

  • Subagent Driven Development

    Asvarox/allkaraoke

    A skill your agent uses when executing implementation plans with independent tasks in the current session

    261 GitHub starsUsed in 37 repos~1.2k tokens
    Agent WorkflowsAuto-check passed
  • Executing Plans

    GanyuanRan/Aegis

    A skill your agent uses when executing a written implementation plan across sessions or with review checkpoints.

    1.3k GitHub starsUsed in 1 repo~2.3k tokens
    Agent WorkflowsAuto-check passed
  • Exec

    umputun/cc-thingz

    Execute plan tasks sequentially using subagents. An agent skill from umputun/cc-thingz.

    483 GitHub stars~8k tokensUpdated 2 days ago
    Agent WorkflowsAuto-check passed
  • Builds an app, site, bot or feature from a spoken idea through requirements, spec, tickets, subagent work, review and acceptance, with a live progress dashboard.

    403 GitHub stars~2.5k tokensUpdated 5 days ago
    Agent WorkflowsAuto-check: notes
  • Workflow Orchestration

    vxcozy/workflow-orchestration

    Disciplined task execution with planning, verification, and self-improvement loops.

    115 GitHub stars~1k tokensUpdated 5 mo ago
    Agent WorkflowsAuto-check passed
  • Blueprint

    byungjunjang/jangpm-meta-skills

    Agentic system design blueprint generator. An agent skill from byungjunjang/jangpm-meta-skills.

    121 GitHub stars~2k tokensUpdated 2 mo ago
    Agent WorkflowsAuto-check passed

More from catlog22/Claude-Code-Workflow

All 82 skills in this repo
  • Prompt Generator

    catlog22/Claude-Code-Workflow

    Generate or convert Claude Code prompt files — command orchestrators, skill files, agent role definitions, or style conversion of existing files.

    2.1k GitHub starsUsed in 1 repo~4.7k tokens
    Auto-check: notes
  • Ccw Help

    catlog22/Claude-Code-Workflow

    CCW command help system. An agent skill from catlog22/Claude-Code-Workflow.

    2.1k GitHub stars~2.5k tokensUpdated 3 mo ago
    Auto-check passed
  • Team Ultra Analyze

    catlog22/Claude-Code-Workflow

    Deep collaborative analysis team skill. An agent skill from catlog22/Claude-Code-Workflow.

    2.1k GitHub starsUsed in 1 repo~2.4k tokens
    Auto-check: notes
  • Brainstorm

    catlog22/Claude-Code-Workflow

    Unified brainstorming skill with dual-mode operation — auto mode (framework generation, parallel multi-role analysis, cross-role synthesis) and single role analysis.

    2.1k GitHub stars~4.8k tokensUpdated 3 mo ago
    Auto-check: notes
  • Ccw Chain

    catlog22/Claude-Code-Workflow

    Chain-based CCW workflow orchestrator. An agent skill from catlog22/Claude-Code-Workflow.

    2.1k GitHub stars~1.1k tokensUpdated 3 mo ago
    Auto-check: notes
  • Delegation Check

    catlog22/Claude-Code-Workflow

    Check workflow delegation prompts against agent role definitions for content separation violations.

    2.1k GitHub stars~2.8k tokensUpdated 3 mo ago
    Auto-check: notes

Categories

Questions about Workflow Plan

What does Workflow Plan do?

Planning pipeline with multi-mode routing (plan/verify/replan). Workflow Plan is an agent skill from catlog22/Claude-Code-Workflow. Planning pipeline with multi-mode routing (plan/verify/replan).

When should I use Workflow Plan?

Workflow Plan fits situations like: tasks that involve Subagents; tasks that involve Task management; tasks that involve Planning.

How do I install Workflow Plan in Claude Code?

Run `npx skills add catlog22/Claude-Code-Workflow --skill workflow-plan -a claude-code`. Or copy the skill folder (.codex/skills/workflow-plan in catlog22/Claude-Code-Workflow) into .claude/skills/workflow-plan in your project. Claude Code loads it when a task matches its description.

How do I install Workflow Plan in Codex?

Run `npx skills add catlog22/Claude-Code-Workflow --skill workflow-plan -a codex`. Or copy the skill folder (.codex/skills/workflow-plan in catlog22/Claude-Code-Workflow) into .agents/skills/workflow-plan in your project. Codex loads it when a task matches its description.

Can I use Workflow Plan in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add catlog22/Claude-Code-Workflow --skill workflow-plan -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/workflow-plan, .gemini/skills/workflow-plan, .github/skills/workflow-plan and .opencode/skills/workflow-plan in your project.

What does Workflow Plan need to run?

SKILL.md names no scripts, command-line tools or credentials: Workflow Plan is instructions for the agent only. Its frontmatter pre-approves these tools: spawn_agent, wait_agent, send_message, followup_task, close_agent, request_user_input, Read, Write, Edit, Bash, Glob, Grep.

Does Workflow Plan access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Workflow Plan safe to install?

Our automated static check of SKILL.md found notes only (pre-approves every shell command (allowed-tools: bash)), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Workflow Plan use?

Workflow Plan is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Workflow Plan use?

About 5.5k tokens (SKILL.md is roughly 22k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Workflow Plan?

Skills that share tags, products or a category with Workflow Plan: Subagent Driven Development (Asvarox/allkaraoke, 261 stars), Executing Plans (GanyuanRan/Aegis, 1.3k stars), Exec (umputun/cc-thingz, 483 stars) and Autopilot End-to-End Builder (nick-vels/skills, 403 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Workflow Plan?

catlog22 (a GitHub user) maintains it in catlog22/Claude-Code-Workflow, which has 2,130 GitHub stars. The repository holds 82 skills in this directory. The repository was last updated on June 18, 2026.

Source: catlog22/Claude-Code-Workflow on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.