By lvlup-sw
Orchestrate a structured, local-first SDLC workflow with durable state, subagent teams, and quality gates — from design exploration and task decomposition to parallel implementation, code review, and PR management.
Run a lightweight oneshot workflow — plan + laddered (tier-scaled) implement + optional PR
Decompose the unified docs/specs/ artifact into verification-laddered tasks
Prune stale workflows from the pipeline (dry-run → confirm → apply)
Start refactor workflow for code improvement
Re-inject workflow state and behavioral guidance into current context
Use this agent when dispatching implementation tasks to a subagent in an isolated worktree — verification scales with the task's risk tier (the verification ladder), not a universal test-first ceremony. <example> Context: Orchestrator is dispatching a task from an implementation plan user: "Implement the agent spec handler (task-003)" assistant: "I'll dispatch the exarchos-implementer agent to implement this task on the verification ladder in an isolated worktree." <commentary> An implementation task at any verification tier triggers the implementer agent. </commentary> </example>
Use this agent when a task has failed and needs diagnosis and repair with adversarial verification. <example> Context: A delegated task failed its quality gates or tests user: "Task-005 failed its test-adequacy gate — fix it" assistant: "I'll dispatch the exarchos-fixer agent to diagnose and repair the failure." <commentary> Failed task requiring root cause analysis and targeted fix triggers the fixer agent. </commentary> </example>
Use this agent when performing read-only code review for quality, design compliance, and test coverage. <example> Context: Feature implementation is complete and needs review user: "Review the agent spec handler for code quality" assistant: "I'll dispatch the exarchos-reviewer agent to analyze code quality and design compliance." <commentary> Code review request triggers the reviewer agent for read-only analysis. </commentary> </example>
Use this agent for low-complexity scaffolding tasks — file creation, boilerplate generation, and structural setup. <example> Context: Orchestrator needs new files or boilerplate created user: "Create the directory structure and stub files for the new feature" assistant: "I'll dispatch the exarchos-scaffolder agent to generate the scaffolding in an isolated worktree." <commentary> Simple file creation and boilerplate generation triggers the scaffolder agent with concise output. </commentary> </example>
Dispatch implementation tasks to agent teammates in git worktrees. Triggers: 'delegate', 'dispatch tasks', 'assign work', or /delegate. Spawns teammates, creates worktrees, monitors progress. Supports --fixes flag. Do NOT use for single-file changes or polish-track refactors.
Collaborative design exploration that authors the Design & Rationale section of the one unified docs/specs/ spec. Triggers: 'brainstorm', 'ideate', 'explore options', or /ideate. Default is a one-pass design preamble; the deep rung gates the 2-3 approach divergent loop and the discover bridge. Do NOT use for task decomposition (that is /plan, which adds Decomposition to the same doc) or code review.
Code improvement workflow with polish and overhaul tracks. Triggers: 'refactor', 'clean up', 'restructure', 'reorganize', or /refactor. Phases: explore, brief, implement, validate. Existing code only — Do NOT use for bug fixes (/debug) or new features (/ideate).
Dispatch implementation tasks to agent teammates in git worktrees. Triggers: 'delegate', 'dispatch tasks', 'assign work', or /delegate. Spawns teammates, creates worktrees, monitors progress. Supports --fixes flag. Do NOT use for single-file changes or polish-track refactors.
Collaborative design exploration that authors the Design & Rationale section of the one unified docs/specs/ spec. Triggers: 'brainstorm', 'ideate', 'explore options', or /ideate. Default is a one-pass design preamble; the deep rung gates the 2-3 approach divergent loop and the discover bridge. Do NOT use for task decomposition (that is /plan, which adds Decomposition to the same doc) or code review.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Your agents forget. Exarchos doesn't.
Persistent SDLC state for any AI coding agent. It survives /clear, auto-compaction, and a blown context window.
First-class with Claude Code, Codex, Cursor, OpenCode, and Copilot; works with any agent that can run a CLI.
Install · What's different · What you get · Architecture · Docs
A local, event-sourced governance engine for coding tasks.
A plan.md per feature. CLAUDE.md rewritten between sessions. Summaries scrawled before /clear so the next session has something to start from. Phases enforced by you, reminding the agent. It works. It's also manual, and one long context window away from the agent ignoring all of it.
/clearCome back to any suspended workflow with /rehydrate.
❯ /exarchos:rehydrate payments-v2-migration
Workflow Rehydrated: payments-v2-migration
Phase: implementing | Type: feature
Task Progress: 4 of 7 complete · last commit on feature/payments-v2
Next Action: Continue task 5 (gates pending). Run /delegate or pick up manually.
The state was never in your conversation. It lives in an append-only event log, and /rehydrate is just a projection that rebuilds the workflow document for a fresh context window. The whole thing fits in about 2,500 tokens.
A state machine owns the phase transitions, not a paragraph in CLAUDE.md. The questions between phases — "is this actually implemented?", "does it match the design?" — run as TypeScript checks against your diff and git history, not as prompts the agent can talk itself out of. You approve the design and you approve the merge. The middle runs on its own.
Run /ideate to start.
The CLI is the universal surface. Each runtime talks to it through whatever invocation it speaks natively.
| Runtime | Transport | Skill rendering | Slash commands |
|---|---|---|---|
| Claude Code | Plugin + MCP | First-class (rendered + hooks) | Yes (/ideate, /plan, etc.) |
| Codex CLI | MCP | First-class | Via Codex's command surface |
| Cursor | MCP | First-class | Via Cursor's MCP integration |
| OpenCode | CLI | First-class | Via OpenCode's runtime |
| GitHub Copilot CLI | CLI | First-class | Via Copilot's runtime |
| Anything else | CLI | Generic bundle | Whatever your agent supports |
The CLI works everywhere. For Claude Code, the plugin is the recommended path.
Standalone CLI / MCP server (any agent, any runtime):
# Unix (macOS / Linux)
curl -fsSL https://lvlup-sw.github.io/exarchos/get-exarchos.sh | bash
# Windows (PowerShell)
irm https://lvlup-sw.github.io/exarchos/get-exarchos.ps1 | iex
exarchos --version
exarchos doctor
exarchos mcp # starts MCP server over stdio
exarchos onboard
One command drives the repo to a green doctor: it detects the runtimes and VCS on your PATH, writes or reconciles agent config, installs the matching skills, registers the SessionStart binding, then verifies. Re-running is safe — it reconciles drift only.
| Flag | Effect |
|---|---|
--new <name> | Scaffold a fresh project in <name>/, then onboard it. |
--dry-run | Print the reconcile plan; write nothing, emit no events. |
--force | Overwrite hand-edited config (preserved otherwise). |
--no-hooks | Skip the SessionStart hook binding. |
--runtime <id> | Target an explicit runtime (claude, codex, opencode, copilot, cursor, generic); bypasses detection. |
To re-check without writing, run exarchos doctor. To re-apply just the remediable diff, run exarchos doctor --fix. If a step fails, onboard exits non-zero and leaves already-applied steps in place — fix the cause and re-run to pick up the residual diff.
/plugin marketplace add lvlup-sw/.github
/plugin install exarchos@lvlup-sw
Backend code quality skills for Claude Code. Nine skills that audit, critique, harden, simplify, humanize, and design-constrain your architecture — the backend half of impeccable.
npx claudepluginhub lvlup-sw/exarchosMulti-agent development workflow (/crew) — planning, implementation, review, documentation, evals, and audit gates. Commands, agents, and skills are auto-discovered from the repo.
Agent orchestration harness for Claude Code — campaign persistence, fleet coordination, intent routing
Corca Workflow Framework — consolidated hooks and skill orchestration for structured development sessions
Implementation planning, execution, and PR creation workflows with multi-agent collaboration
Development toolkit for Claude Code — plan, implement, ship, review, and assess features with AI-assisted workflows. Progressive zero-config init: auto-configures with sensible defaults on first skill invocation, no upfront ceremony required. Three-tier ceremony model: swift (lightweight), standard (mid-ceremony spec-plan-execute), and thorough (full pipeline) with severity-aware scope routing. Five entry points: arn-planning (scope router, spec, plan), arn-implementing (execute plans, swift, or standard changes), arn-shipping (commit, push, PR), arn-reviewing-pr (PR feedback), arn-assessing (codebase health). Includes arn-code-sketch for UI preview, arn-code-swift for quick implementations, and arn-code-standard for mid-ceremony changes. Includes arn-code-catch-up for retroactive documentation of out-of-pipeline commits. Pipeline preference persistence for streamlined repeat sessions. Batch pipeline: arn-code-batch-planning (multi-feature planning), arn-code-batch-implement (parallel worktree execution), arn-code-batch-merge (conflict-aware merge), arn-code-batch-simplify (cross-feature quality).
AI-powered development workflow automation - Phase-based planning, implementation orchestration, preflight code quality checks with security scanning, ship-it workflow, and development principles generator for CLAUDE.md