Skip to content

commands.md

Latest commit

 

History

History
366 lines (261 loc) · 14.9 KB

commands.md

File metadata and controls

366 lines (261 loc) · 14.9 KB
Raw

Public Slash Command Reference

Public installed workflow commands for Spec-Flow.

This page covers the slash-command or prompt surface installed into a consumer project. It does not cover the packaged installer CLI (npx spec-flow ...) and it does not count maintainer-only source-repo commands.

Quick Navigation

Core Commands

Core orchestration commands for feature development workflows.

Command Description
/feature [description|slug|continue|next|epic:<name>|epic:<name>:sprint:<num>|sprint:<num>] Execute feature development workflow from specification through production deployment with automated quality gates and manual approval checkpoints
/quick <description> Implement small bug fixes and features (<100 LOC) without full workflow. Use for single-file changes, bug fixes, refactors, and minor enhancements that can be completed in under 30 minutes.
/help [verbose] Analyze workflow state and provide context-aware guidance with visual progress indicators and recommended next steps
/route-agent [task description] Analyze task and route to specialist agent (backend, frontend, database, qa, debugger, or reviewer) using scoring algorithm and shared routing rules

Usage Example:

# Start a new feature
/feature "Add dark mode toggle to user settings"

# Quick fix
/quick "Fix typo in welcome message"

# Resume current feature
/feature continue

# Get workflow guidance
/help

Phase Commands

Individual workflow phases - can be run standalone or via /feature orchestration.

Command Description
/spec <feature-description> [--interactive] [--yes] [--skip-clarify] Generate complete feature specification with research, requirements analysis, and quality gates
/clarify [spec-identifier or empty for auto-detect] Reduce spec ambiguity via targeted questions with adaptive auto-invocation (planning is 80% of success)
/plan [feature-name or epic-slug] Generate implementation plan from spec using research-driven design (meta-prompting for epics)
/tasks [--ui-first] Generate TDD task breakdown from plan.md with test-first sequencing and mockup-first mode (--ui-first)
/validate [feature-slug] [--quick|--constitution] Analyze spec, plan, and tasks for consistency violations, breaking changes, and constitution compliance. Generates analysis-report.md with CRITICAL/MAJOR/MINOR findings.
/implement [feature-slug] Execute all implementation tasks from tasks.md with test-driven development, parallel batching, and atomic commits
/optimize [feature-slug or empty for auto-detection] Run parallel quality gates (performance, security, accessibility, code review, migrations, Docker) and block deployment on failures
/debug [feature-slug] [options] Execute the systematic debugging workflow, track failures in error-log.md, and generate session reports
/finalize Finalize documentation (CHANGELOG, README, help docs), update GitHub milestones/releases, and cleanup branches after production deployment

Usage Example:

# Create specification
/spec "User authentication with OAuth 2.1"

# Clarify ambiguities
/clarify

# Generate implementation plan
/plan

# Break down into tasks (with UI-first mockups)
/tasks --ui-first

# Validate consistency
/validate

# Implement
/implement

# Run quality gates
/optimize

# Finalize after deployment
/finalize

Epic Commands

Multi-sprint workflows for complex features (>16 hours, multiple subsystems).

Command Description
/epic [epic description | slug | continue | next] Execute multi-sprint epic workflow from interactive scoping through deployment with parallel sprint execution and self-improvement
/implement-epic [epic-slug] Execute multiple sprints in parallel based on dependency graph from sprint-plan.xml for epic workflows

Usage Example:

# Start an epic
/epic "User authentication with OAuth 2.1"

# Resume epic
/epic continue

# Implement epic sprints in parallel
/implement-epic auth-system

When to use /epic vs /feature:

  • Use /epic for: >16 hours, multiple subsystems, 2-5 sprints, >5 API endpoints, >3 database tables
  • Use /feature for: ≤16 hours, single subsystem, 1 sprint, ≤5 endpoints, ≤3 tables

Deployment Commands

Staging and production deployment workflows.

Command Description
/ship [continue|status] Deploy feature through automated staging validation to production with rollback testing
/ship-staging [feature-slug] Create PR to main with auto-merge, triggering staging deployment via CI/CD pipeline with health checks and deployment metadata capture
/validate-staging Validate staging deployment before production through automated test review, rollback capability testing, and guided manual validation
/ship-prod [--version major|minor|patch] Automated staging→production promotion via git tag creation, triggering GitHub Actions deployment, and creating GitHub release
/deploy-prod Deploy feature directly to production by triggering GitHub Actions workflow, monitoring deployment, extracting rollback IDs, and verifying health
/deploy-status Display comprehensive deployment workflow status showing current phase, completed tasks, quality gates, and deployment information
/deployment-budget Display Vercel and Railway deployment quota usage with 24h rolling window analysis, quota reset predictions, and deployment strategy recommendations
/validate-deploy [staging|production] Validate deployment readiness by simulating production builds, Docker images, and migrations locally to catch failures before consuming CI/CD quota

Usage Example:

# Deploy to staging (staging-prod model)
/ship-staging

# Validate staging deployment
/validate-staging

# Promote to production
/ship-prod

# Or direct to production (direct-prod model)
/deploy-prod

# Check deployment status
/deploy-status

# Check deployment quota
/deployment-budget

Build Commands

Local builds and branch management.

Command Description
/build-local Build and validate locally for projects without remote deployment (prototypes, experiments, local-only dev)
/branch-enforce [--fix] [--verbose] [--json] [--current-branch-only] [--max-hours N] [--warn-hours N] [--default-branch NAME] [--no-color] Audit branches for age violations and enforce trunk-based development (warn 18h, block 24h unless feature-flagged)

Usage Example:

# Build locally (local-only model)
/build-local

# Check branch age
/branch-enforce

# Fix stale branches
/branch-enforce --fix

Project Commands

Project initialization, roadmap management, and configuration.

Command Description
/init-project ["project-name"] [--with-design] [--update|--force|--write-missing-only] [--config FILE] [--ci] [--non-interactive] Generate 8 project design documents (overview, architecture, tech-stack, data, API, capacity, deployment, workflow) via interactive questionnaire or config file
/init-brand-tokens Generate OKLCH design tokens with WCAG validation and Tailwind v4 integration by scanning existing code or creating new palette
/roadmap [add|brainstorm|move|delete|search|list|milestone|epic] [additional args] Manage product roadmap via GitHub Issues (brainstorm, prioritize, track). Auto-validates features against project vision (from overview.md) before adding to roadmap.
/constitution <action>: <description> Add, update, or remove engineering principles in docs/project/engineering-principles.md with atomic versioned commits
/update-project-config <configuration change description> Update project configuration settings (deployment model, scale tier, quick changes policy) with atomic commits

Usage Example:

# Initialize project docs
/init-project

# Initialize with design system
/init-project --with-design

# Initialize brand tokens
/init-brand-tokens

# Manage roadmap
/roadmap add "User authentication"
/roadmap brainstorm
/roadmap list

# Update constitution
/constitution "add: A11Y - Accessibility | policy=WCAG 2.1 AA minimum"

# Update project config
/update-project-config "Change deployment model to staging-prod"

Quality Commands

Quality gates and CI/CD integration.

Command Description
/fix-ci [pr-number] Diagnose and fix CI/deployment blockers for pull requests to enable safe deployment
/gate-ci [--epic <epic-name>] [--verbose] Run CI quality gate checks (tests, linters, type checks, coverage ≥80%) before epic state transition
/gate-sec Run security quality gate (SAST, secrets detection, dependency scanning) to ensure no HIGH/CRITICAL security issues before deployment

Usage Example:

# Fix CI failures
/fix-ci 123

# Run CI quality gate
/gate-ci

# Run security gate
/gate-sec

Infrastructure Commands

Workflow health monitoring and self-improvement.

Command Description
/audit-workflow Analyze workflow effectiveness and generate improvement recommendations after epic completion
/heal-workflow Apply workflow improvements discovered during audit with approval workflow
/workflow-health Display workflow health dashboard with velocity trends, quality metrics, and improvement tracking

Usage Example:

# Audit workflow after epic
/audit-workflow

# Apply improvements
/heal-workflow

# View health dashboard
/workflow-health
/workflow-health --detailed
/workflow-health --trends

Meta Commands

Developer tools for creating commands, skills, agents, and prompts.

Command Description
/add-to-todos <todo-description> Add todo item to TO-DOS.md with context from conversation
/check-todos List outstanding todos and select one to work on
/create-agent-skill [skill description or requirements] Create or edit Claude Code skills with expert guidance on structure and best practices
/create-hook Invoke create-hooks skill for expert guidance on Claude Code hook development
/create-meta-prompt [task description] Create optimized prompts for Claude-to-Claude pipelines (research → plan → implement)
/create-prompt [task description] Expert prompt engineer that creates optimized, XML-structured prompts with intelligent depth selection
/create-slash-command [command description or requirements] Create a new slash command following best practices and patterns
/create-subagent [agent idea or description] Create specialized Claude Code subagents with expert guidance
/audit-skill <skill-path> Audit skill for YAML compliance, pure XML structure, progressive disclosure, and best practices
/audit-slash-command <command-path> Audit slash command file for YAML, arguments, dynamic context, tool restrictions, and content quality
/audit-subagent <subagent-path> Audit subagent configuration for role definition, prompt quality, tool selection, XML structure compliance, and effectiveness
/heal-skill [optional: specific issue to fix] Heal skill documentation by applying corrections discovered during execution with approval workflow
/whats-next Analyze the current conversation and create a handoff document for continuing this work in a fresh context
/run-prompt <prompt-number(s)-or-name> [--parallel|--sequential] Delegate one or more prompts to fresh sub-task contexts with parallel or sequential execution
/anti-duplication [optional: what you want to implement] Search codebase for existing patterns before implementing new code to prevent duplication
/breaking-change-detector [optional: change description] Detect breaking API/schema/interface changes before implementation and suggest safe migration paths
/caching-strategy [optional: operation to cache] Apply intelligent caching to avoid redundant work and speed up workflow execution by 20-40%
/hallucination-detector [optional: technical suggestion to validate] Detect and prevent hallucinated technical decisions by validating against project's documented tech stack
/parallel-optimize [optional: context or phase to optimize] Apply parallel execution optimization to detect and parallelize independent operations for 3-5x speedup
/resolve-dependencies [package manager command, e.g., "npm install react@18"] Detect and resolve package dependency conflicts before installation (npm, pip, cargo, composer)
/sync-task-status [optional task operation, e.g., "mark T001 complete"] Enforce atomic task status updates through task-tracker (prevent manual edits to NOTES.md/tasks.md)
/enforce-git-commits [optional: phase/task context] Enforce git commits after phases/tasks with auto-commit and safety checks
/announce-release [optional: version number] Post release announcement to X (Twitter) with GitHub link reply

Usage Example:

# Create development artifacts
/create-slash-command "New command for XYZ"
/create-agent-skill "Skill for ABC"
/create-subagent "Agent for DEF"

# Audit artifacts
/audit-skill .claude/skills/my-skill/SKILL.md
/audit-slash-command .claude/commands/my-command.md

# Prevent issues
/anti-duplication "user authentication"
/breaking-change-detector "Change API endpoint /v1/users to /v2/users"
/hallucination-detector "Use React Query for data fetching"

# Optimize workflow
/parallel-optimize
/caching-strategy

# Task management
/add-to-todos "Investigate performance bottleneck"
/check-todos

Maintainer-Only Commands

Maintainer-only source-repo commands are intentionally excluded from this public reference. Document them in explicit maintainer docs instead of counting them as public operator surface.

Getting Help

In Claude Code:

  • Type /help to see context-aware guidance for your current workflow state
  • Type /help verbose for detailed state information

Documentation:

Support:

Last updated: 2026-04-23