By TerrysPOV
Manage Claude Code daemon agents with persistent memory, scheduled cron jobs, heartbeat monitoring, and Telegram/Discord integration. Scaffold new agents, create custom skills, enforce shell command guards, and control daemon lifecycle.
Create a new ClaudeClaw agent (identity, workflow, personality, scheduled jobs)
Update an existing ClaudeClaw agent (workflow, personality, jobs, discord, or delete)
Clear session and start fresh
View or modify heartbeat settings
Show Discord bot status and manage global session
Use when the user wants to create a new agent teammate, scaffold a scheduled Claude persona, or asks to "create an agent", "new agent", "add an agent", "add a job", "scheduled task", "scaffold an agent", "build an agent", "make an agent", or "/claudeclaw:create-agent". Trigger phrases include "I want an agent that", "create a teammate", "new agent persona", "scheduled claude agent".
Create new skills for Claude Code. Use when users ask to create a skill, add a skill, make a new command, build a skill, add a slash command, create a plugin skill, or define a new automation. Trigger phrases include "create a skill", "new skill", "add a skill", "make a command", "build a skill", "I want a skill that", "add slash command", "create automation".
Search and install skills from skills.sh and GitHub repos. Use when users ask to find skills, install skills, download skills, add skills from GitHub, search for skills, browse skills, get a skill, or want new capabilities. Trigger phrases include "install skill", "find skill", "search skills", "add skill", "download skill", "get skill from github", "skills.sh", "browse skills", "what skills are available", "I need a skill for".
Add Telegram reaction directives in assistant replies. Use when the user asks for Telegram reaction tags, react directives, emoji reactions, or wants reply text to include [react:<emoji>] metadata. Trigger phrases include "telegram react", "reaction directive", "add [react]", "emoji reaction", "react tag", and "telegram reply format".
Use when the user wants to modify an existing agent — change personality, workflow, add or edit a scheduled job, update Discord channels, swap a job's model, or delete the agent. Trigger phrases include "update an agent", "edit agent", "modify agent", "change agent", "/claudeclaw:update-agent", "add a job to <agent>", "remove job from <agent>", "delete agent <name>".
Admin access level
Server config contains admin-level keywords
Executes bash commands
Hook triggers when Bash tool is used
npx claudepluginhub terryspov/claudeclaw-plus --plugin claudeclaw-plusBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
ClaudeClaw, plus the heavy stuff. Governance, orchestration, persistent memory, hardened web UI.
ClaudeClaw+ is a sister project to moazbuilds/claudeclaw — the lightweight Claude Code daemon you already know. Plus exists to house the features that are too heavy or too opinionated for the core repo: a full governance and policy layer, durable multi-step orchestration, persistent cross-session memory, and a hardened web UI.
Everything from upstream lives here too. Plus has now diverged significantly from upstream, so alignment is maintained on a best-effort manual basis rather than via an automated sync.
Note: Please don't use ClaudeClaw+ for hacking any bank system or doing anything illegal. Same rules apply.
ClaudeClaw+ is built on top of the original moazbuilds/claudeclaw, created by @moazbuilds. The core daemon, Telegram/Discord adapters, heartbeat, web dashboard, skills system — all of that comes from upstream and the amazing contributors who built it.
Upstream contributors — thank you:
Plus has diverged significantly from moazbuilds/claudeclaw, so the previous automated daily sync has been retired. Alignment is now best-effort and manual: upstream fixes and features are reviewed and ported by hand when they make sense for Plus.
There is no guaranteed turnaround on upstream changes landing here. If you need a specific upstream commit applied to Plus, open an issue.
| Category | ClaudeClaw | ClaudeClaw+ | OpenClaw |
|---|---|---|---|
| Anthropic Will Come After You | No | No | Yes |
| API Overhead | Directly uses your Claude Code subscription | Same | Nightmare |
| Setup & Installation | ~5 minutes | ~5 minutes | Nightmare |
| Isolation | Folder-based and isolated as needed | Folder-based + per-agent scope | Nightmare |
| Reliability | Simple reliable system | Simple + durable workflows that survive restarts | Nightmare |
| Feature Scope | Lightweight features you actually use | Everything in ClaudeClaw, plus governance and orchestration | 600k+ LOC nightmare |
| Security | Average Claude Code usage | Policy engine, audit log, CSRF protection | Nightmare |
| Cost Control | Manual | Automatic token budgets + model routing per task type | Nightmare |
| Memory | Claude internal memory + CLAUDE.md | Persistent cross-session MEMORY.md with dual-layer write guarantee | Nightmare |
| Multi-step Jobs | Cron + heartbeat | DAG orchestrator with dependency resolution and resumable state | Nightmare |
As of v2.0 (May 2026), ClaudeClaw+ runs on a new event-bus architecture by default. The claude -p subprocess path (runtime: "pty") remains a permanent first-class option — especially for enterprise deployments where API billing is the safer or only viable route (audit trails, cost ceilings, regulatory constraints). Operators pick per deployment: bus for subscription-billed interactive work, claude -p for API-billed programmatic work. Both are fully supported, side by side.
Here's what changes when you pick the bus runtime:
Cron-like daemon that runs Claude prompts on a schedule
Self-evolving Claude Code daemon — Telegram + Discord bridges, cron jobs, learned skills, SQLite state
Turn Claude Code into a persistent agent — memory, personality, voice, messaging, and more.
Autonomous task execution daemon that monitors Claude usage windows and executes predefined tasks automatically. Keep Claude working 24/7 on your projects.
Context management and multi-agent orchestration with performance optimization tools
High-intelligence Claude Code copilot with deep code reasoning, evidence-driven planning, orchestration-first execution, model routing, context budgeting, CI/CD integration, enterprise security, plugin development, prompt engineering, performance profiling, agent teams, channels (event-driven autonomy with CI webhook, mobile approval relay, Discord bridge, and fakechat dev profile), interactive tutorials, LSP integration, security-hardened hook script library, MCP Prompts coverage, common workflow packs, runtime selection guide, computer-use patterns, checkpointing, scheduled-task blueprints, repo bootstrap scanner, hook policy engine (8 installable packs), layered memory deployment, role-based subagent packs (implementer, debugger, migration-lead, dependency-auditor, release-coordinator), 5 agent-team topology kits, autonomy operating mode (4 profiles + 3 gates), and a queryable 15-tool MCP documentation server with autonomy advisor.