By kangig94
Orchestrate multi-agent AI workflows with structured planning, debugging, code review, and test generation across Claude and Codex environments. Use agents for architecture analysis, requirements gap detection, and adversarial test generation. Automate pipeline steps with delegate support for cross-host execution and deterministic orchestration.
Architecture & debugging advisor. Use PROACTIVELY when reviewing code structure, design patterns, dependency analysis, or debugging complex issues. Also participates as a structural reviewer in the /plan protocol. NOT for requirements analysis (gap-finder).
Plan & code change critic. Use PROACTIVELY when reviewing implementation plans, schema changes, or significant code modifications. NOT for code analysis (architect) or requirements gathering (gap-finder).
Systematic bug diagnosis via hypothesis testing, reproduction tracing, and root cause analysis.
Requirements gap analyst. Catches missing questions, undefined guardrails, scope risks, and edge cases before planning. Use PROACTIVELY when scoping new features, API changes, state lifecycle changes, or concurrency behavior modifications. NOT for project scanning (scanner), code debugging (debugger), or plan review (critic).
Generate a diverse, differentiated discussion persona based on role and topic. Spawned in parallel by discuss-lead.
Use when deep investigation is needed — project structure, requirement gaps, or root cause diagnosis. Supports --delegate.
Submit a bid or speech in an active --user discuss session
Use when encountering a bug, error, or unexpected behavior that needs diagnosis and fix.
Use when code needs simplification — recently modified code by default, or a specified scope.
Use when a topic benefits from multiple perspectives debating before a decision.
Matches all tools
Hooks run on every tool call, not just specific ones
Executes bash commands
Hook triggers when Bash tool is used
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.
Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
Claude Code already knows how to code. Coral teaches it how you work.
Coral is a CLI-first plugin backed by a persistent local coordinator for orchestration, sessions, discussion, and knowledge-base workflows.
Requirements: Node.js 22+
# Claude Code:
/plugin marketplace add https://github.com/kangig94/coral
/plugin install coral
# Codex (also enables --delegate cross-model delegation):
npm install -g @openai/codex
codex plugin marketplace add kangig94/coral
# Restart Codex, then run /plugins and install Coral from the Coral marketplace.
# Update the Codex marketplace and installed plugin cache:
codex plugin marketplace upgrade coral
Run this on any existing project:
/coral:analyze what does this codebase do?
https://github.com/user-attachments/assets/881f1a14-9f4f-4d3d-8023-59610eb13ac4
/coral:init-project
Coral scans your stack and generates .claude/ — conventions, review agents, architecture docs — tailored to your project.
Generated agents aren't boilerplate — they encode evaluation rubrics calibrated to your project's stack and audience. Claude follows your rules, not generic defaults.
/coral:init-project # existing project
/coral:init-project "React + FastAPI" # tech stack hint
/coral:init-project "multi-tenant SaaS REST API with Go" # full description
my-project/
+ .claude/
+ CLAUDE.md ← project hub: build commands, workflow, critical rules
+ agents/
+ code-critic.md ← code quality review
+ ... ← domain agents (React, Go, ML, infra, etc.)
+ rules/
+ conventions.md ← naming, git, style
+ ... ← domain rules, auto-activated by file path
+ docs/
+ architecture.md ← module map, dependency graph
Browse this repository's .claude/ folder for a real example.
pathfind → preplan → plan → ralph
(explore) (define) (design) (implement + verify)
# Know the problem, need a plan:
/coral:plan add retry logic to the API client
# Have symptoms, not sure what's wrong:
/coral:pathfind API is slow, DB hits limits, users are complaining
# Complex problem, need alignment first:
/coral:preplan race condition in the session manager
# Have a plan, just implement it:
/coral:ralph implement the caching layer
# Bug — diagnose, plan, fix in one shot:
/coral:bugfix why does session lookup return null?
Each stage produces an artifact that feeds the next. Enter at any point — skip stages you don't need.
pathfind — "I have problems but don't know what to build." Clusters symptoms, investigates root causes (spawns scanner for codebase analysis), generates divergent directions through orthogonal lanes, and spawns pioneer for elegant alternatives. Outputs a ranked direction list with scoring matrix. Hands off the chosen direction to preplan.
preplan — "I know the direction but need agreement on scope."
Fills a 7-item agreement (problem statement, success criteria, scope, assumptions, affected systems, constraints, approach direction)
autonomously from codebase analysis, then presents to the user for correction.
Spawns pioneer to find elegant alternatives for uncertain items, offering default/minimal/elegant spectrums.
Produces pre-{topic}.md — the contract that plan must satisfy.
plan — "I need a design before I build."
Multi-round review loop: dispatches architect + critic + resolver as a workflow,
synthesizes findings, edits the plan file, and evaluates an exit gate (round=N sets the round budget).
The resolver classifies findings (Adopt/Adapt/Defer/Diverge), applies changes,
and decides whether another round is needed based on finding severity and nature.
Produces {topic}.md with acceptance criteria, implementation phases, and execution order (dependency graph + parallel batches).
ralph — "I have a plan (or prompt). Just build it."
Persistent executor with verification loop.
In plan mode, reads the execution order and dispatches batches — parallelizing independent ACs.
Every completion claim requires fresh verification evidence (lint → build → test).
--red spawns a red-attacker in parallel to write adversarial tests targeting blind spots.
--team uses Agent Teams for parallel AC execution.
# Deeper iteration — up to 3 review rounds per phase:
/coral:plan round=3 add retry logic to the API client
# Adversarial testing — spawns a red-attacker to target blind spots:
/coral:ralph --red implement the caching layer
npx claudepluginhub kangig94/coral --plugin coralClaude harness - A harness for solo developers (Vibecoders) to handle full-cycle contract development.
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Superpowers Plus core skills library for Claude Code: planning, execution routing, TDD, debugging, and collaboration workflows
Persistent file-based planning for AI coding agents. Crash-proof markdown plans (task_plan.md, findings.md, progress.md) that survive context loss and /clear, with an opt-in completion gate and multi-agent shared state. Manus-style. Works with Claude Code, Codex CLI, Cursor, Kiro, OpenCode and 60+ agents via the SKILL.md standard. Includes Arabic, German, Spanish, and Chinese (Simplified and Traditional).
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques