The request asks what to use for coding agents, with no platform or constraint stated. Cursor is an AI coding agent that hands tasks off to cloud agents 111, Claude Code-style delegation and background coding agents are offered by OpenClaw.Direct 414, and OpenHands is an open-source platform to build and run coding agents 9. Other candidates include jcode 6, fx 3, HAR 10 and Robota 2.
- 1
- 2
- 3
- 4
OpenClaw.Directopenclaw.directverified and signed
A managed hosting service for an AI assistant with a coding-agent skill that delegates coding work to Codex, Claude Code or OpenCode as background workers, and an MCP server 414.
official:MCP server · OpenClaw Direct
for agents: claude-agent-sdk · coding-agent · tmux-agents
Sources
- 1AI Coding Agent for Building Ambitious Software | Cursorcursor.com · read 2026-10-29
- 2Robota | TypeScript AI agent SDK and coding CLIrobota.io · read 2026-10-29
- 3fx - Tiny, open, native coding agentfx.sh · read 2026-10-29
- 4Fully Managed OpenClaw Hosting — AI Assistant — OpenClaw.Directopenclaw.direct · read 2026-10-29
- 6jcode | Coding agentsjcode.sh · read 2026-10-29
- 9OpenHands | Open Source AI Coding Agent Platformopenhands.dev · read 2026-10-29
- 10HAR — Open agent harness (CLI + MCP) for coding agentsharproject.dev · read 2026-10-29
- 11cli-for-agents: declares "cli-for-agents" (the record we read and signed)unfragile.ai · read 2026-10-24
- 14claude-agent-sdk: declares "claude-agent-sdk" (the record we read and signed)unfragile.ai · read 2026-10-24
For agents: the 10 actions behind this answer
cli-for-agents · cli-for-agents verified and signed
Designs or reviews CLIs so coding agents can run them reliably: non-interactive flags, layered --help with examples, stdin/pipelines, fast actionable errors, idempotency, dry-run, and predictable structure. Use when building a CLI, adding commands, writing --help, or when the user mentions agents, t
· Robota website read, tools not yet checked
Build AI agents with the Robota TypeScript SDK, or use the coding CLI built from the same public components. Available in public beta.
· fx website read, tools not yet checked
Tiny, open, native coding agent. A minimal CLI written in Zig, optimized for research and embeddability as part of larger systems.
claude-agent-sdk · claude-agent-sdk verified and signed
Build autonomous AI agents with Claude Agent SDK. Structured outputs guarantee JSON schema validation, with plugins system and hooks for event-driven workflows. Prevents 14 documented errors. Use when: building coding agents, SRE systems, security auditors, or troubleshooting CLI not found, structu
· Typeship website read, tools not yet checked
Make your API agent-friendly. Typeship turns a Swagger, OpenAPI, or GraphQL spec into a CLI, an MCP server, and native SDKs in TypeScript, Python, and Go, then keeps them current with reviewed pull requests.
· jcode website read, tools not yet checked
jcode is a high performance, open-source coding agent, written in Rust. Agent swarms, background tasks, and a fraction of the RAM, installed with one command.
· agent-device website read, tools not yet checked
Open-source device layer for coding agents: control mobile, web, TV and desktop apps on devices and simulators, local or in the cloud, via CLI, MCP or Node.js.
agents-md · agents-md verified and signed
Creates and maintains concise AGENTS.md and CLAUDE.md project instruction files. Use when asked to create AGENTS.md, update AGENTS.md, maintain agent docs, set up CLAUDE.md, document repository agent conventions, or keep coding-agent instructions minimal and reference-backed.
ssh · ssh verified and signed
Establish and manage SSH connections to remote machines, including key generation, configuration, and file transfers. Use when connecting to remote servers, executing remote commands, or transferring files via SCP.
· Harproject website read, tools not yet checked
HAR is an open-source agent harness — CLI and MCP server — for multi-agent coding workflows. Isolated worktrees, deterministic verification, and software-factory observability for Claude Code, Cursor, and Codex.
Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.