raw-cli MCP Server
lploc94/raw-cli
[](https://agentrank-ai.com/tool/lploc94--raw-cli/?utm_source=badge&utm_medium=readme&utm_campaign=agentrank_badge) The agent harness you can fork. A working agent loop with compaction, sessions, approvals, a terminal REPL, a local dashboard and ACP, where every tool, skill, hook and policy is a file you can change. Build your own internal or coding agent on any model.
claude mcp add agentrank -- npx -y agentrank-mcp-server Overview
lploc94/raw-cli is a TypeScript tool licensed under MIT. The agent harness you can fork. A working agent loop with compaction, sessions, approvals, a terminal REPL, a local dashboard and ACP, where every tool, skill, hook and policy is a file you can change. Build your own internal or coding agent on any model. Topics: agent-client-protocol, ai-agent, cli, coding-agent, llm, mcp, typescript.
Ranked #5273 out of 24846 indexed tools.
Actively maintained with commits in the last week.
Ecosystem
Score Breakdown
105 stars → growing community
Last commit 7d ago → actively maintained
No issues filed → no history to score
1 contributor → solo project
No dependents → no downstream usage
Weights: Freshness 20% · Issue Health 20% · Dependents 22% · Stars 10% · Contributors 8% · How we score →
How to Improve
Badge all embed codes →
[](https://agentrank-ai.com/tool/lploc94--raw-cli/?utm_source=badge&utm_medium=readme&utm_campaign=agentrank_badge) <a href="https://agentrank-ai.com/tool/lploc94--raw-cli/?utm_source=badge&utm_medium=readme&utm_campaign=agentrank_badge"><img src="https://agentrank-ai.com/api/v1/badge/lploc94--raw-cli" alt="AgentRank"></a> Matched Queries
From the README
# raw-cli `raw` is the agent harness you can fork. The parts every agent needs already work: the tool loop, context compaction, saved sessions, approvals, a terminal REPL, a local browser dashboard and the Agent Client Protocol for editors. Everything an agent does is a file you can change, so you build an internal agent, a coding agent or an ops agent by editing parts instead of starting from an empty loop. - **Fork a tool.** Built-in tools are plain plugins. Copy one with Fork to local, or write your own as one `index.mjs`. - **Teach a procedure.** Skills are Markdown files the agent lists and loads when it needs them. - **Set the rules.** Ordered policy rules allow, ask or deny a tool or a pattern, and hooks run your scripts on session, prompt and tool events. - **Connect your systems.** Selected MCP tools and runtime variables sit next to the built-ins. - **Share the result.** Package an agent with its tools and skills as a `.rawpkg` that your team installs. Each agent selecRead full README on GitHub →
Get the weekly AgentRank digest
Top movers, new tools, ecosystem insights — straight to your inbox.