Best use case
devloop is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Start development workflow - lightweight entry point
Teams using devloop should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/devloop/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How devloop Compares
| Feature / Agent | devloop | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Start development workflow - lightweight entry point
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Devloop - Smart Development Workflow
The main entry point for devloop. Detects your current state and suggests the most relevant actions. **You do the work directly.**
## Step 1: Detect State
Run the state detection script:
```bash
"${CLAUDE_PLUGIN_ROOT}/scripts/check-devloop-state.sh"
```
Parse the JSON output:
- `state`: Current state identifier
- `priority`: 1-7 (lower = more urgent)
- `details`: State-specific information
- `suggestions`: Recommended actions
## Step 2: Display Status and Options
Based on detected state, display status and present contextual options:
### State: `not_setup` (Priority 1)
```
Devloop not set up in this project.
The .devloop/ directory doesn't exist yet.
```
```yaml
AskUserQuestion:
questions:
- question: "What would you like to do?"
header: "Setup"
multiSelect: false
options:
- label: "Create a plan (Recommended)"
description: "Autonomous exploration -> actionable plan"
- label: "Deep exploration"
description: "Comprehensive investigation with spike report"
- label: "Quick task"
description: "Small, well-defined fix or change"
- label: "GitHub issues"
description: "Start from an existing issue"
```
**Routing:**
- "Create a plan" -> `/devloop:plan $ARGUMENTS`
- "Deep exploration" -> `/devloop:plan --deep $ARGUMENTS`
- "Quick task" -> `/devloop:plan --quick $ARGUMENTS`
- "GitHub issues" -> `/devloop:issues`
### State: `active_plan` (Priority 2)
Display plan status from details:
```
Active Plan: [plan_title]
Progress: [done]/[total] tasks ([pending] remaining)
Next: [next_task]
```
```yaml
AskUserQuestion:
questions:
- question: "What would you like to do?"
header: "Action"
multiSelect: false
options:
- label: "Continue working (Recommended)"
description: "Pick up where you left off"
- label: "Ship progress"
description: "Commit and optionally create PR"
- label: "View full plan"
description: "Review the complete plan"
- label: "Start fresh"
description: "Archive current plan and start new"
```
**Routing:**
- "Continue working" -> `/devloop:run`
- "Ship progress" -> `/devloop:ship`
- "View full plan" -> Display `.devloop/plan.md` content
- "Start fresh" -> Archive then ask what to do
### State: `uncommitted` (Priority 3)
```
Uncommitted Changes Detected
[total_changes] files changed ([staged] staged, [unstaged] unstaged)
```
```yaml
AskUserQuestion:
questions:
- question: "You have uncommitted changes. What would you like to do?"
header: "Git"
multiSelect: false
options:
- label: "Commit changes (Recommended)"
description: "Create a commit with current changes"
- label: "Review changes"
description: "See what's changed before committing"
- label: "Continue without committing"
description: "Start new work, keep changes staged"
- label: "Stash changes"
description: "Save changes for later"
```
**Routing:**
- "Commit changes" -> `/devloop:ship`
- "Review changes" -> Run `git diff --stat` then `git diff`
- "Continue without committing" -> Proceed with $ARGUMENTS or ask
- "Stash changes" -> Run `git stash push -m "devloop: work in progress"`
### State: `open_bugs` (Priority 4)
```
Open Bugs: [bug_count]
```
```yaml
AskUserQuestion:
questions:
- question: "[bug_count] open bug(s) found. What would you like to do?"
header: "Bugs"
multiSelect: false
options:
- label: "Fix a bug"
description: "View and work on open bugs"
- label: "Start new feature"
description: "Work on something new instead"
- label: "Deep exploration"
description: "Research or explore an idea"
```
**Routing:**
- "Fix a bug" -> List bugs from GitHub issues then create plan
- "Start new feature" -> Ask for details, `/devloop:plan`
- "Deep exploration" -> `/devloop:plan --deep`
### State: `backlog` (Priority 5)
```
Backlog: [feature_count] feature(s) waiting
```
```yaml
AskUserQuestion:
questions:
- question: "[feature_count] item(s) in backlog. What would you like to do?"
header: "Backlog"
multiSelect: false
options:
- label: "Work on backlog item"
description: "Pick from existing features/tasks"
- label: "Deep exploration"
description: "Research or explore an idea"
- label: "Quick task"
description: "Small, well-defined fix"
```
**Routing:**
- "Work on backlog item" -> List items then `/devloop:plan --from-issue N`
- "Deep exploration" -> `/devloop:plan --deep`
- "Quick task" -> `/devloop:plan --quick`
### State: `complete_plan` (Priority 6)
```
Plan Complete: [plan_title]
All [total] tasks finished!
```
```yaml
AskUserQuestion:
questions:
- question: "Plan is complete. What would you like to do?"
header: "Complete"
multiSelect: false
options:
- label: "Ship it (Recommended)"
description: "Commit and optionally create PR"
- label: "Archive and start new"
description: "Move to archive, begin fresh"
- label: "Review before shipping"
description: "Look over the work one more time"
```
**Routing:**
- "Ship it" -> `/devloop:ship`
- "Archive and start new" -> Archive then ask what's next
- "Review before shipping" -> `/devloop:review`
### State: `clean` (Priority 7)
```
Ready for new work!
No active plans or pending changes.
```
```yaml
AskUserQuestion:
questions:
- question: "What would you like to work on?"
header: "Start"
multiSelect: false
options:
- label: "Create a plan (Recommended)"
description: "Autonomous exploration -> actionable plan"
- label: "Deep exploration"
description: "Comprehensive investigation with spike report"
- label: "GitHub issues"
description: "View and work from GitHub issues"
- label: "Quick task"
description: "Small, well-defined fix"
```
**Routing:**
- "Create a plan" -> `/devloop:plan`
- "Deep exploration" -> `/devloop:plan --deep`
- "GitHub issues" -> `/devloop:issues`
- "Quick task" -> `/devloop:plan --quick`
## Step 3: Handle Arguments
If `$ARGUMENTS` is provided and non-empty:
- Skip state display
- Use arguments as task description
- Route appropriately based on state:
- `active_plan` -> Ask: continue plan or start new with this task?
- Other states -> Create new plan with the description
## Key Principles
1. **You (Claude) do the work** - Don't spawn subagents for tasks you can do yourself
2. **Skills on demand** - Load with `Skill: skill-name` only when needed
3. **Minimal questions** - One question at a time, not multi-part interrogations
4. **Fast iteration** - Ship working code, then improve
## Workflow Commands
| Command | Purpose |
|---------|---------|
| `/devloop` | Smart entry point (this command) |
| `/devloop:plan` | Autonomous exploration -> actionable plan |
| `/devloop:plan --deep` | Comprehensive exploration with spike report |
| `/devloop:plan --quick` | Fast path for small tasks |
| `/devloop:plan --from-issue N` | Start from GitHub issue |
| `/devloop:run` | Execute plan autonomously |
| `/devloop:epic` | Create multi-phase epic with TDD |
| `/devloop:run-epic` | Execute epic phase-by-phase |
| `/devloop:fresh` | Save state and exit cleanly |
| `/devloop:new` | Create GitHub issue |
| `/devloop:issues` | List GitHub issues |
| `/devloop:review` | Code review |
| `/devloop:ship` | Commit and/or PR |
## Files
- `.devloop/plan.md` - Current task plan (or promoted epic phase)
- `.devloop/epic.md` - Multi-phase epic plan
- `.devloop/local.md` - Project settings (git workflow, etc.)
- `.devloop/next-action.json` - Fresh start state
- `.devloop/spikes/` - Spike reports (from --deep exploration)
---
**Now**: Detect state and present options.Related Skills
devloop-audit
Audit devloop against Claude Code updates to identify integration opportunities. Use after Claude Code releases, monthly maintenance, or when exploring new features.
Example Skill
Brief description of what this skill does and the domain expertise it provides.
vulnerability-patterns
Index of vulnerability detection pattern skills. Routes to core patterns (universal) and language-specific patterns for security scanning.
vuln-patterns-languages
Language-specific vulnerability detection patterns for JavaScript/TypeScript, Python, Go, Java, Ruby, and PHP. Provides regex patterns and grep commands for common security vulnerabilities.
vuln-patterns-core
Universal vulnerability detection patterns applicable across all programming languages. Includes hardcoded secrets, SQL/command injection, path traversal, and configuration file patterns.
scan
Run a security assessment using deterministic static analysis tools with LLM-powered triage
results
View the most recent security scan results without re-running the scan
remediation-library
Index of security remediation skills. Routes to specialized skills for injection, cryptography, authentication, and configuration vulnerabilities.
remediation-injection
Security fix patterns for injection vulnerabilities (SQL, Command, XSS). Provides language-specific code examples showing vulnerable and secure implementations.
remediation-crypto
Security fix patterns for cryptographic vulnerabilities (weak algorithms, insecure randomness, TLS issues). Provides language-specific secure implementations.
remediation-config
Security fix patterns for configuration and deployment vulnerabilities (path traversal, debug mode, security headers). Provides language-specific secure implementations.
remediation-auth
Security fix patterns for authentication and authorization vulnerabilities (credentials, JWT, deserialization, access control). Provides language-specific secure implementations.