Agent-harness picks, decision guides, setup templates, and skills; live GitHub/HN discovery.
Model Context Protocol (MCP) Server: io.github.RyanAlberts/agent-harnesses
The io.github.RyanAlberts/agent-harnesses MCP server provides access to a curated collection of AI agent harnesses, orchestration frameworks, and harness techniques. It centers on selecting and guiding infrastructure, including pick_infrastructure for live GitHub/HN discovery.
🛠️ Key Features
“Pick and decision guides” via agent-harness selection
pick_infrastructure with live GitHub/HN discovery
Topics focused on agent harnesses and tooling under MCP
🚀 Use Cases
Querying a “best of” list of agent harnesses and harness techniques
Finding infrastructure choices using live GitHub/HN signals
Supporting reliable agentic systems via harness patterns
⚡ Developer Benefits
Curated resources for agent-framework and agent-harness engineering
Structured focus on agentic-ai, tools, and LLM/LLM-agents terminology
MCP integration for agents to query the list
⚠️ Limitations
Source material only describes selection and live discovery; it does not document additional MCP tools or configuration details.
📰 This week: new harnesses and the biggest star gains, every Sunday. To get it as a GitHub notification, click Watch, then Custom, then Releases.
🧰 Templates and Playbooks: copy-paste setup files and step-by-step guides for the harnesses in this list.
🧩 Skills: 10 installable skills that check your own agent setup, including what it can reach, what it claims, and what it spends.
🤖 Agents can query this list — an MCP server (recommend, pick_harness, …), llms.txt & JSON, so your agent recommends harnesses too.
What is an agent harness?
A model answers; an agent acts. An agent harness is the runtime that turns one into the other: the model thinks, the harness decides what that thinking is allowed to touch.
Simon Willison's definition of the agent itself is the cleanest: "an LLM agent runs tools in a loop to achieve a goal." The harness is everything around that loop: which tools exist, what needs approval, what the model sees each turn, what survives a crash. Andrej Karpathy named the architecture back in 2023: the model is "the kernel process of a new Operating System", and the harness is the rest of that OS, its scheduler, permissions, and memory. The SWE-agent paper proved the stakes by coining the agent-computer interface: how tools and feedback are presented changes what a model can do, independent of the model. The field's advice has since converged on investing here rather than in framework plumbing, from Anthropic's build-simple guidance to Jerry Liu's argument that the framework era is over and the layers that matter now are skills, tools, and context quality. Those are the layers this list catalogs.
Why harnesses matter
Better models make harnesses more important: more capabilities mean more failure modes, and production needs retry logic, fallbacks, and validation. Harness quality, not just model quality, determines whether agents actually ship. This list ranks projects by relevance to harness concerns (environment, orchestration, lifecycle, guardrails) and by stars/activity.
The benchmark data now backs this up. On SWE-bench Pro, "swapping the agent harness changed pass@1 more than many model upgrades do" (AINews, Aug 8 2026, citing analysis by @joelniklaus). Same model, different harness: 23% to 52% pass@1 on GLM-5.2, and 15% to 36% on Gemma 4 26B. Harness rankings barely transfer across models (rank correlation -0.05), so a small model in the right harness can approach a much larger model in the wrong one.
The gap is widest on long-horizon work. A bare frontier model was verified at about 30% on ARC-AGI-3; Prime Agent's harness took Opus 5 to 95.5%, and the YC Paper Club talk on why the harness matters more than the model (September 2026) walks through how. The gains come from a new class of self-improving harnesses, not from piling on scaffolding: the best of them stay thin and expose only what the model cannot do for itself. And because rankings barely transfer across models, the harness choice is a pairing that must be re-asked whenever the model changes. Who says so, what they measured, and what the claim does not mean is its own page.
That is the problem the MCP server in this repo solves. Point your agent at it and it can call recommend or pick_harness to choose a harness matched to your model and task, instead of inheriting whichever harness someone else benchmarked.
The landscape at a glance
Every project in the list, plotted by adoption surface area (the simplicity ↔ capability axis) against GitHub stars. Colors are categories; the largest projects in each tier are labeled.
The same projects placed by how much unsupervised rope they're designed to give (autonomy) and what happens when a run dies (recovery). In the tables below, ★ marks headless-ready projects and ✱ marks durable ones. Both charts regenerate from the list data on every refresh.
How to Pick a Harness
Start with the guide, then the head-to-head decision pages — grounded in the same data as the tables below:
How to pick a harness: six questions that turn this list into a decision, plus the chart to internalize first (the harness moves scores more than the model)
How to test-drive a harness: the two-week trial protocol, with a fair setup, tasks from your own repos, seven measurements, and the walk-away test
OpenClaw vs Hermes — the always-on personal-agent debate: presence vs discipline, plus what the field reports actually say
Managed vs self-hosted always-on agents: Grok Bot, Claude Managed Agents, QM, OpenClaw, Hermes, and OpenJarvis; who owns the computer and who pays for idle time
The list tells you which harness to use. These tell you how to set it up. Templates are files you copy into your project; Playbooks walk you through one task, step by step.
Templates
One AGENTS.md for every coding agent: One briefing file that Codex, Claude Code, Cursor, OpenCode, GitHub Copilot, Gemini CLI, and Aider all read, so you write your build commands, conventions, and hard rules once instead of once per tool.
Safe Claude Code settings: permissions and a guard hook: Project settings that stop Claude Code from force-pushing, wiping work, reading secrets, or piping downloads into a shell, while leaving everyday commands alone. Two files, copy and commit.
Minimal agent harness in Python: A working coding-agent harness in about 180 lines of Python: the loop, three tools, a permission gate, a context file, a turn budget, and a transcript you can resume after a crash. Copy it to learn how harnesses work, or as the start of your own.
Playbooks
Build your own agent harness: Build a working coding agent in about an hour: a loop, three tools, permissions, a context file, a budget, and crash recovery, in one Python file you fully understand. You finish with the minimal harness template running on your own repo.
Write one AGENTS.md for every coding agent: Write one briefing file that Codex, Claude Code, Cursor, OpenCode, Copilot, Gemini CLI, and Aider all read, test that each tool actually loaded it, and keep it short enough to help instead of hurt. You finish with the AGENTS.md template filled in for your repo.
Skills
A skill is a folder your agent loads when a task calls for it. Each one below runs a script on your own setup and hands back a result you can check: what your agent can reach, which of its claims had no proof, where your tokens went. They run on your machine, and anything that changes a file or spends money waits for your yes.
Pick a harness
Test-drive coding agents on tasks from your own repo: Runs Claude Code, Codex, and Gemini CLI on tasks taken from your own git history, scores each one with your repository's tests, and reports how many tasks each passed, the dollars per pass, the minutes, and the size of each change.
Set it up
Check which instruction files your coding agents load: See which instruction files Claude Code, Codex, Gemini CLI, OpenCode, Cursor, GitHub Copilot, and Aider load from your repo, what each one cuts or skips, and whether the commands in those files still work.
Grade your MCP tools the way a model reads them: Point it at an MCP server or a saved tool list to grade each tool's purpose, parameters, annotations, and token size, or let it find the MCP servers configured across your coding agents and count what they load into every session.
Keep it safe
Test your agent's guardrails against dangerous commands: Checks whether the permission rules and hooks you already have in Claude Code, Codex, Gemini CLI, OpenCode, or Cursor stop about 90 dangerous commands, shows each one that gets through with a tested fix, and counts how often your rules interrupt real work.
Stop a runaway agent session: A hook for Claude Code and Codex that stops a live session when the agent repeats the same call, keeps failing, or spends past a dollar cap you set.
Check what your coding agent can reach: A one-minute probe that runs through your agent's own shell and shows which secret files, trust files, Docker, and sudo access it really has, then compares that with what your sandbox settings claim.
Check its work
Check your agent's 'tests pass' claims: Find out how often your coding agent said tests passed without a passing run to back it up, and whether the current change quietly weakened your tests. It reads your local session logs and sends nothing anywhere.
Turn the rules your agent breaks into hooks: Finds the rules in AGENTS.md, CLAUDE.md, and GEMINI.md that your coding agent breaks, counts every break in your recent sessions, and turns each broken rule into a hook (a check your agent runs before each tool call) that blocks it, tested against those real breaks.
See what it costs and what changed
Find the update where your coding agent started working differently: Splits your own Claude Code or Codex sessions by version, model, or week, measures how the agent works in each, and places the change at the update, or within the span of versions, where the numbers moved, along with anything else that changed then.
See where your coding agent wastes tokens and money: A report on your own Claude Code, Codex, Gemini CLI, and OpenCode sessions that shows which habits cost the most, such as re-read files, huge tool outputs, polling, and breaks that expire the prompt cache (the provider's cheap copy of the conversation), and which failures repeat, each with a fix.
Reader's index: pick by what you want to do, not by category. Tag chips (e.g. mcp · memory) next to each row let you cross-filter by capability — see TAGS.md for the full cross-reference.
This list is also published in machine-readable form, so coding agents and research agents can recommend harnesses — not just humans browsing GitHub:
harnesses.json — every project with category, complexity tier, capability tags, stars, license signal, and a concrete example link, plus the full use-case index.
llms.txt — the entire list in one agent-readable file. Point any agent at the raw URL.
MCP server — recommend (one opinionated pick + alternatives + what to avoid, e.g. repos flagged for star manipulation), compare/compare_for (2–4 harnesses side by side — by id or by task — who leads on which axis incl. researched sandboxing/memory/hooks/prompt-optimization ratings, graveyard warnings, the matching decision guide), pick_harness (ranked, with complexity/autonomy/recovery filters), pick_infrastructure (picks at any level of the infra stack plus a live GitHub/Hacker News discovery pass, so answers aren't limited to this list), search_harnesses, get_harness, list_categories, whats_new (the weekly digest), plus list_comparisons/get_comparison for the decision guides and list_templates/get_template/list_playbooks/get_playbook/list_skills/get_skill so your agent can install a template or a skill for you. Published to PyPI and the official MCP registry as io.github.RyanAlberts/agent-harnesses. One-line install (needs uv):
sh
claude mcp add agent-harnesses -- uvx agent-harnesses-mcp
Or hire a skeleton
Don't just read the list — agents/ ships three agent skeletons: open-source agents that run on the AI subscription you already pay for. Clone the file, customize the instructions, done. All three work against the current week's data and deliver to Slack or Notion when either is connected:
harness-scout — describe what you're building; it picks your harness, with evidence and a graveyard check.
stack-auditor — flags the harnesses in your codebase that died, and can trace your agent session logs to show how the harness steers your technical decisions.
harness-radar — weekly movement briefing: climbers, arrivals, deaths, graduations.
⭐ Stars — GitHub star count, captured 2026-11-15; tables sort by stars descending.
⚖️ Simplicity ↔ capability — adoption surface, 4 tiers: super simple (a format, one concept) → mostly simple (thin layer) → slightly complex (real SDK) → complex (product suite).
★ Headless-ready — designed for unattended runs, batches, and fleets (the top of the autonomy scale: step-gated → checkpoint-gated → bounded → headless).
✱ Durable — persisted execution state survives restarts mid-task (the top of the recovery scale: none → retry → resumable → durable).
archived upstream — the maintainers archived the repo, so it gets no new fixes. Kept in the table for reference, never in the use-case picks.
✅ Open source — ✅ standard OSS license · ⚠️ source-available/restricted · ❓ no or unclear license.
🏷️ Tags — capability chips auto-derived from descriptions; full cross-reference in TAGS.md.
🎯 Examples — one concrete "show me it in action" link per project, not a docs root.
Every project's full autonomy and recovery tier is plotted in the grid above and carried in harnesses.json and llms.txt; scores are editorial, from public docs — maintainer corrections via issue/PR are merged fast.
Formats, runtimes, and patterns that reveal context, tools, or instructions in layers—index first, details on demand—to control tokens and improve agent focus (the "map, not encyclopedia" principle).
Compresses tool outputs, logs, files, and RAG chunks with content-aware compressors before they reach the model—claimed 20% fewer tokens for coding agents and 60–95% fewer for JSON, same answers. Ships as a library, HTTP proxy, or MCP server, so it drops in front of whatever harness you already run. mcp · rag
Curated .cursorrules and skills that leverage Cursor's index-then-load model; the canonical collection for rules-as-progressive-disclosure in the IDE. ide
Context-window optimization layer that sandboxes tool output before it reaches the model (claimed 98% reduction) and persists session memory across 17 agent platforms via MCP and hooks—progressive disclosure applied to tool results, not just instructions. mcp · memory · sandbox
Open format for repo-scoped agent briefings; nested AGENTS.md files scope instructions per directory, so agents get a map of what exists and load only what's relevant. Read by 20+ tools including Codex, Cursor, and Copilot. ide · typescript
Build LangGraph agents with large tool sets; retrieval and on-demand tool loading so agents scale beyond context without stuffing every schema upfront. tool-discovery · python
Active tool discovery for autonomous agents: model requests tools by requirement; hierarchical semantic routing over 308 servers / 2,797 tools with ~98% token reduction (APIBank). tool-discovery
ICLR 2025: unified tool retrieval and calling via generation; 47k+ tools without context stuffing—retrieval and invocation in one generative step. tool-discovery · python
Semantic tool retrieval for LLMs; serves only the tools the user query demands (MCP-compatible), unlimited tool sets with zero context penalty. mcp · tool-discovery
Turnkey coding agents you install and run: IDE extensions, terminal CLIs, Dockerized workspaces. Each entry notes which part is the harness (the agent loop, tool wiring, approval model) versus the UI shell (VS Code extension, TUI, browser client).
DeepSeek's official open-source agent harness (dsh), built so that everything is a plugin: the model adapter, tool registry, session log, sandbox and approval policy, and the agent loop itself are Cordis plugins you replace from configuration, and one plugin tree boots as a Web UI, a headless runner, a desktop app, an SDK server, or an ACP server. Developer preview, with breaking changes expected. browser · sandbox · typescript
✅
slightly complex (plugin-tree runtime, Web UI, headless, SDK)
Open-source terminal coding agent (formerly sst/opencode; transferred to anomalyco). The harness is a multi-provider tool-call loop (Claude, OpenAI, Gemini, local) with strong plugin and MCP support; the TUI is the shell. 100% OSS, very actively shipped. mcp · provider-agnostic · cli · tui · typescript
OpenAI's terminal coding agent. The harness is the sandboxed tool-call loop with multi-provider support; the CLI is the shell. Reference implementation for "official CLI that ships code." sandbox · provider-agnostic · cli
The upstream AI agent toolkit behind this list's oh-my-pi fork: a unified multi-provider LLM API, agent loop, and TUI shell providing the harness that oh-my-pi's Rust rewrite builds on. provider-agnostic · tui · rust
Google's first-party terminal agent for Gemini. The harness is the plugin/MCP tool-call loop; the terminal is the shell—Google's parallel to Claude Code / Codex, not just an API. mcp · cli · typescript
Dockerized software-engineering agent. The harness is the bash/editor/browser toolset with micro-agents and event-stream session bridging; Docker is the sandbox. Main OSS choice for teams self-hosting autonomous repo work. memory · browser · sandbox · python
VS Code extension whose harness is a plan-then-act loop with per-step human approval and cost transparency; the VS Code integration is the UI shell. Open-source counterweight to Cursor. ide · typescript
Lightweight terminal coding agent oriented to open models (DeepSeek, Kimi, Qwen). The harness is a code-execution loop — the model writes code, the harness executes it with confirmation gates; the CLI is the shell. The original "let the LLM run code on my machine" project, reborn for open weights. cli · python
Block-originated Rust agent, now stewarded by the Linux Foundation's Agentic AI Foundation (aaif-goose/goose). The harness is the MCP/ACP extension model with recipes and provider choice; there's no fixed UI slot—you bolt it into whatever shell you use. mcp · rust
Open-source Rust terminal coding agent for any model (40+ providers, local models included). The harness pairs a goal-driven work loop with permission modes (Ask, Auto-Review by a reviewer model, Full Access) and an optional workspace-write sandbox, and /receipts lists every file it changed, command it ran, and web or MCP call it made; the TUI is the shell. mcp · sandbox · cli · tui · rust
✅
slightly complex (Rust CLI, any model, permission modes, MCP)
Terminal coding agent (fork of Pi) that wires the IDE into the harness: hash-anchored edits, a 32-tool loop tuned per-model, LSP rename/references/diagnostics on every write, a real DAP debugger (lldb/dlv/debugpy), long-lived Python + Bun execution kernels that call back into the agent's tools, browser control, and 40+ providers (Claude/OpenAI/Gemini/local). ~55k-line Rust core. browser · provider-agnostic · cli · ide · rust
Alibaba's official terminal coding agent, forked from Gemini CLI's agent loop and retuned for Qwen models. The harness is the same sandboxed tool-call loop as its upstream; the terminal is the shell. sandbox · cli · typescript
Charm's terminal coding agent (Charm's fork of the original OpenCode). The harness is the tool-calling loop with session persistence; the Bubble Tea TUI is the shell. memory · cli · tui
Kanban-style fleet manager for running Claude Code, Codex, or any coding agent across many tasks at once. The harness contribution is the task-queue/review layer on top of whichever agent executes; not an agent loop itself.
OpenAI's harness for fanning a task out into many isolated, autonomous coding-agent implementation runs and surfacing the ones that pass, so a team manages outcomes instead of supervising each session. sandbox
VS Code extension and CLI in the Cline/Roo-Code lineage — a natural pick now that Roo-Code is archived upstream. The harness is an approval-gated autonomous-mode loop with a provider/tool marketplace; the IDE is the shell. mcp · cli · ide · typescript
SpaceXAI's terminal coding agent. The harness is a Rust agent runtime that edits files, runs shell commands, searches the web, and keeps workspace checkpoints; it runs in a full-screen TUI, headless for scripts and CI, or inside editors over the Agent Client Protocol. The public repo is synced from the company monorepo. browser · cli · tui · rust
✅
slightly complex (official CLI, Rust runtime, headless and ACP)
VS Code/Cursor extension in the Cline lineage. The harness is the approval-gated agent with custom modes and a strong MCP story; the IDE is the UI. Popular community fork when you want that workflow without the upstream extension. mcp · workflow · ide · typescript
Prime Intellect's self-improving coding and research agent. The harness is a recursive language model loop (context as a variable, sub-agents as function calls inside a persistent IPython REPL) plus a Continual Harness that lets the agent create, read, update, and delete its own prompts, skills, memory, and sub-agents mid-run; daemon-backed sessions survive a closed terminal, and autonomous mode runs on turn, token, and time budgets. 95.5% on ARC-AGI-3 with Opus 5; the terminal UI is the shell. memory · multi-agent · cli · typescript
Open-source desktop harness positioned as a local, free alternative to Claude Cowork and Codex: multi-agent workspace orchestration in a self-hosted app rather than a hosted product. multi-agent · local
Local-first desktop workspace harness for Claude Code and other agents: multi-agent sessions, Git worktrees, code diffs, a skill marketplace, and chat-app access (WeChat, Telegram, WhatsApp). memory · multi-agent · typescript
Coding agent for the desktop, the terminal, and CI. The harness builds a repo map before each task (tree-sitter symbols, import graph, git co-change), edits by symbol through AST and LSP, then runs typecheck, lint, and tests in the same turn. This repo holds the public SoulForge v2 core under BUSL-1.1; the current Empryo v3 app is developed privately under its own license. workflow · cli · typescript
⚠️ BUSL-1.1
slightly complex (repo map, AST/LSP symbol edits, any model)
TypeScript coding-agent runtime shared by a terminal TUI, a desktop app, and a VS Code extension. The harness keeps each task's goal, scope, and success criteria in a task contract outside the chat history, blocks a claim of done until tests, a diff, or a verification command back it, and takes a git checkpoint before each turn's first edit. The main README is Chinese; an English README is linked. cli · ide · tui · typescript
Runs multiple coding agents in parallel, each in its own sandboxed VM, locally or in the cloud, from one command. The harness contribution is the VM-per-agent isolation and fleet fan-out layer; whichever agent runs inside owns the loop. sandbox · typescript
Open-source AI IDE for Claude Code, Codex, OpenCode, and more. The harness contribution is the workspace/session orchestration layer: run multiple coding agents in parallel, locally or in the cloud, with isolated workspaces, reusable workflows, and shared team context. multi-agent · sandbox · ide · typescript
Rust coding-agent runtime that began as a reimplementation of Pi and now develops on its own. The embeddable harness owns a provider-agnostic streaming agent loop, tool execution, JSONL sessions that resume after a crash from the last committed frame, and a native plugin ABI; the terminal TUI is a shell over the same crates, and a deterministic offline provider covers tests. memory · provider-agnostic · cli · tui · rust
Self-hosted runner that takes a GitHub issue labeled ai-ready to a reviewed, merged pull request and a tagged release: each run gets its own branch and worktree, Pi writes the change, a separate review session checks it against the issue's acceptance criteria, and only the reviewed head merges. The harness contribution is the issue-to-release delivery and review gate on top of Pi; it does not own an agent loop. python
Command-line orchestrator for coding agents (drives Pi and Codex subagents): every task gets a dedicated branch/worktree and a typed lifecycle — read-only preflight, risk-scaled merge-queue review, release-readiness gates — with receipt-backed changes. The harness contribution is the git-native task/merge orchestration layer on top of whichever agent executes; not an agent loop itself. multi-agent · typescript
Skill packs, slash-command libraries, meta-prompting frameworks, and official SDKs that give you the harness (the agent loop, planning, memory, hooks) without bundling a specific IDE or CLI shell.
Performance-oriented harness pack for Claude Code and 13 other harnesses (Codex, Cursor, OpenCode, Gemini CLI, more): skills, instincts, memory, security, research-first workflows. Treats harness engineering itself as the performance lever. memory · cli · ide
Anthropic's official Agent Skills repository: SKILL.md-based folders (instructions, scripts, resources) Claude dynamically loads on Claude Code, Claude.ai, and the API. The reference for progressive-disclosure skill packs in 2026.
Garry Tan's Claude Code skill stack: 23 slash-command modes (CEO/eng/design review, QA, ship, browse, retro, …) that structure one assistant as a virtual engineering team. Daily driver while running YC. typescript
Addy Osmani's production-grade skill pack: 24 engineering skills and 4 specialist agent personas that encode senior-dev workflows (spec through deploy) across 70+ coding agents including Claude Code, Cursor, and Copilot. The harness contribution is the skill/workflow layer, not a new agent loop. workflow · ide
Large community-curated index of Claude Code skills, slash commands, status lines, and plugins—resources for extending the harness, not a harness itself, but the most-followed catalog of the genre.
Cross-harness marketplace of drop-in subagents and skills for Claude Code, Codex CLI, Cursor, OpenCode, and Copilot; specialized, production-ready agent definitions you install rather than hand-write. multi-agent · cli · ide
Skill for persistent, file-based planning across long-running coding-agent sessions: crash-proof markdown plans, session recovery after /clear/compaction, and a deterministic completion gate—Manus-style planning as a drop-in harness layer via the Agent Skills standard. memory
LM-driven harness built for SWE-bench: edit state, command execution, and issue-focused loop—the reference agent stack next to the benchmark itself. memory · evals · python
Goal-backward planning and wave-based execution over fresh context windows; avoids context rot by design. Python/JS meta-prompting for Claude Code, OpenCode, Gemini CLI. cli · python
Spec-first layer that wraps an existing coding agent (Claude Code, Codex CLI, OpenCode, and ten more runtime families) instead of replacing it. The harness pins an acceptance spec before the run and checks it afterwards itself, keeps verify commands and expected outputs out of the worker's contract so the agent is asked for the outcome rather than handed the test, and runs its evolution loop unattended, with lineage kept in an event store that survives restarts. mcp · cli · python
Google's official CLI and skill pack that layers agent-creation, evaluation, and deployment skills on top of whatever coding assistant you already run, rather than shipping its own agent loop—the harness as a config/skills add-on, not a new runtime. evals · cli
iFlytek's self-hosted registry for publishing, versioning, and governing agent skill packages—the harness config layer treated as an enterprise artifact store rather than a CLI or IDE shell. local · cli · ide
Reference implementation from the Meta-Harness paper: an academic testbed for harness-engineering research, not a product—useful as a citation-grade baseline rather than something you'd run in production.
Repo-scoped research harness: builds function-call and module-dependency graphs to explore only what's needed; large relative gains on MLE-bench and GitTaskBench with lower token use. workflow · python
Config layer that chains LLM councils for planning, Ralph loops for iterative refinement, and OpenCode worktrees for shipping. The harness contribution is the council → loop → worktree pipeline; OpenCode underneath executes. typescript
Claude Code config for AI product managers: CLAUDE.md plus skills for competitive analysis, PRD-from-signal, metric frameworks, stakeholder briefs, and agent eval design. "GStack for PMs." evals
Always-on, self-hosted agents you run as a daemon and talk to from chat apps: gateway runtimes, second brains, and self-improving assistants. The agent as a product you operate, not a library you build with.
Self-hosted, always-on personal agent (formerly Clawdbot/Moltbot): a gateway + event-loop runtime that treats messages, heartbeats, crons, and webhooks as one input queue, persists state to local files, and lives in your chat apps (WhatsApp, Telegram, Slack, Discord). 13,700+ community skills; the fastest-growing repo in GitHub history. typescript · multi-agent
Nous Research's self-improving agent: a learning loop turns experience into reusable skills, builds a persistent user model across sessions, and checkpoints state to disk with rollback; lean enough for a $5 VPS, driven from chat, and model-agnostic (Nous Portal, OpenRouter, OpenAI, or any endpoint). memory · python · provider-agnostic
Self-hosted "AI second brain" harness: chat with your documents, run built-in agent skills (web search, code execution, browsing), and manage multi-user workspaces with a bundled vector DB, so there is no separate RAG stack to wire up. rag · typescript
Ultra-lightweight, self-hosted personal agent framework: the harness is a Python daemon wiring tools, memory, and MCP into chat/webhook front ends (Telegram, Discord, web); minimal footprint alternative to heavier personal-runtime stacks. mcp · memory · local · python
Self-hostable "AI second brain": answers over your docs and the web, custom agents, scheduled automations, and multi-client reach (web, Obsidian, Emacs, WhatsApp). A personal-agent harness with retrieval at the core. python
Open "agentic operating system" (elizaOS): persistent multi-agent runtime with character files, a plugin ecosystem, and social/platform integrations — the harness behind a large share of autonomous social agents. memory · multi-agent · typescript
Organic, prompt-defined personal agent framework: hierarchical sub-agents, persistent memory, browser and code tools, and self-modifying behavior; runs in Docker with a web UI. memory · multi-agent · browser · sandbox · python
❓
slightly complex (prompt-defined, Docker + web UI)
Open agent harness with a built-in personal agent ("Ohmo") that runs across Feishu, Slack, Telegram, and Discord; core tool-use, skills, memory, multi-agent coordination with auto-compaction for multi-day sessions. memory · multi-agent
Y Combinator's multiplayer agent harness for work, open-sourced from months of internal use: every person and room gets scoped memory, files, credentials, permissions, crons, web apps, and a durable sandbox; sessions, memory, and the work queue live in Postgres, sandboxes are a resource the agent reaches for rather than the place it lives, and Pi, OpenCode, Codex, or Claude Code can drive the same core. Slack and the web app are the shells. memory · sandbox · typescript
Stanford Hazy Research's local-first personal AI harness: on-device model inference (Ollama built in), agent execution, memory, and learning as shared primitives, with on-demand, scheduled, and continuous agents and evaluations that count energy, latency, and dollars alongside accuracy; a cloud model can tune the local configuration once so run time stays on your own hardware. memory · local · python
✅
slightly complex (local-first runtime, scheduled and continuous agents)
Self-modifying personal agent that keeps its identity, durable memory, and history across tasks and restarts. The harness owns the agent loop, tool execution, and specialist-subagent coordination, and it changes its own code only through review gates, with restart checks and rollback kept in git history; the desktop app and headless CLI are interfaces to the same runtime. memory · multi-agent · browser · cli · python
Multi-platform personal agent living in Telegram, Discord, Teams, and the terminal. The harness is a pluggable-backend loop (Claude, Kilo, OpenCode, Codex, OpenAI Agents) with full MCP tool access and persistent background agents (Goals, Heartbeat, Dream); the chat apps are shells. mcp · memory · cli · typescript
Fair-code workflow engine with 400+ nodes and native AI nodes; the self-hosted Zapier that actually does agents and LangChain. workflow · local · typescript
The original autonomous loop: goal in, agent iterates with tools and memory; Forge is the dev framework, Benchmark the eval harness. memory · evals · python
One-stop LLM app platform: visual workflows, RAG pipeline, 50+ tools, model management; "ship from prototype to prod" in a single UI. low-code · rag · python
Python web-agent harness: natural-language goals become browser actions, driven directly over the Chrome DevTools Protocol (it dropped Playwright in August 2025). The biggest community in the browser-agent category. browser · python
Drag-and-drop LangChain UI; deploy flows without code. The low-code sibling to Langflow, with a different component and hosting story. low-code · typescript
Python agents with memory, knowledge bases, tools, and structured outputs; continues the PhiData-era product line under the Agno name—production apps, evals, and pipelines. memory · evals · python
Open-source orchestration framework for context-engineered LLM apps: modular pipelines and agent workflows with explicit control over retrieval, routing, memory, and generation—closer to LangChain's territory than a coding-agent harness. memory · rag · python
Browserbase's SDK for browser agents: natural-language actions (act, extract, observe) and deterministic Playwright code mix in one script, so agent flexibility and repeatable automation live in the same harness. browser · typescript
Google's official Agent Development Kit: code-first Python toolkit for building, evaluating, and deploying agents. Optimized for Gemini but model-agnostic; deploys to Cloud Run / Vertex AI; ships a dev UI with eval and a code-execution sandbox. evals · sandbox · python
✅
complex (official Google SDK, eval, deploy — product suite)
RAG-first: hybrid search, knowledge graphs, multimodal; the framework for "production RAG" when you care more about retrieval than chat UI. vision · rag · workflow · python
AWS-originated orchestrator (now under 2FastLabs): intent classification, streaming, SupervisorAgent; "agent-as-tools" so one agent delegates to a squad. multi-agent
Task-solving and simulation envs for multi-LLM agents; deploy many agents in custom environments without building infra from scratch. multi-agent · python
Tencent Cloud's agent framework: a minimal tool-calling harness designed to perform well with open-source models, positioned as a lighter alternative to heavier orchestration frameworks.
Lightweight multi-agent orchestrator with an AGI-angle; minimal surface, docs-first, for teams that want orchestration without the kitchen sink. multi-agent · python
The "AI software company" multi-agent framework: role-played PM, architect, and engineer agents turn a one-line requirement into specs, designs, and code along an SOP assembly line. The landmark of the genre; development pace has slowed in 2026. multi-agent · python
Conversable agents and group chats; code execution and human-in-the-loop; Microsoft origin, AG2 ecosystem. ⚠️ In maintenance mode since late 2025: no new features, and Microsoft directs new users to Agent Framework. multi-agent · python
Open, invite-free general agent from the MetaGPT team: planning plus tool use over a multi-agent loop, aimed at reproducing Manus-style autonomous task completion on your own keys. multi-agent · python
Microsoft's convergence of AutoGen and Semantic Kernel: build, orchestrate, and deploy agents and multi-agent workflows in Python and .NET, with graph-based workflows and checkpointing — the designated successor harness for both lines. multi-agent · workflow · python
Open-source meta-harness: orchestrates Claude Code, Codex, Cursor, Pi, and custom agents behind one policy/sandboxing layer so teams swap harnesses without rewriting workflows. sandbox · ide · python
Self-hosted multi-agent harness aimed at production workloads: human-in-the-loop checkpoints and a self-improving agent loop, distinct from single-session coding-agent shells. multi-agent · python
Local-first desktop AI team. The harness is a Commander agent that turns one chat goal into a plan and runs specialist agents in parallel or in sequence, with tool approvals, session checkpoints, tool-round budgets and idle watchdogs for stalled runs, durable per-agent memory, and support for built-in or external agent runtimes. memory · multi-agent · typescript
Self-hosted TypeScript multi-agent runtime. The harness is a coordinator that plans a task DAG from one goal at runtime and a deterministic scheduler that runs specialized agents, with durable approvals that pause before a plan, task, or tool call, PostgreSQL-backed checkpoints for restart recovery, and a verifiable run journal for replay; runTasks() keeps the graph fixed when the topology must not change. multi-agent · low-code · workflow · typescript
EverMind's host agent, with its loop split into memory, planning, capability, and action modules. The harness plans complex tasks as DAGs and orchestrates built-in research, coding, design, and on-call agents plus third-party agents (Claude Code, Codex, OpenClaw, Hermes Agent) over ACP, CLI, or OpenAI-compatible APIs, with long-term memory; labeled pre-alpha, with tagged releases since v0.1.0. mcp · memory · multi-agent · cli · python
AG2 (formerly AutoGen): the community-governed continuation of the original AutoGen project after Microsoft's fork diverged—conversable multi-agent groups, code execution, and human-in-the-loop under an open-source AgentOS banner. Graduated off the radar this cycle. multi-agent · python
Self-hostable Community Edition of a unified interface for complete agent harnesses. The harness contribution is the layer around them: each task runs a full harness (Codex, Claude Code, Hermes, Pi, DeepSeek Harness, and others) in its own on-demand sandbox, the harness is a request parameter behind an OpenAI Responses-compatible API (the open Unified Harness Protocol), and the layer handles sessions, streaming, files, and traces. memory · sandbox · python
✅
complex (sandbox-per-run router across harnesses, product suite)
Desktop app that runs a team of coding agents (Claude Code, Codex, OpenCode, Cursor, Copilot, and others) on a kanban board: agents take tasks, message each other, and review each other's work before it is accepted. The harness contribution is the team-coordination layer over existing agent CLIs, with per-action approval or full autonomy; the kanban board is the shell. multi-agent · ide · typescript
Open-source enterprise agent harness: a code-act loop in which the model writes Python that calls MCP, OpenAPI, and LangChain tools as functions inside a restricted sandbox, with results kept as variables, pre-execution verification, reflection, and tool shortlisting; CugaSupervisor coordinates multi-agent teams locally or over A2A, and runtime-enforced policies (intent guards, tool approvals, ToolGuard checks) gate what agents may do. Previously ranked #1 on the AppWorld and WebArena leaderboards; runs on open-weight or hosted models, including air-gapped deployments. Maintained by IBM and open to community contributions. mcp · multi-agent · sandbox · python
✅
mostly simple (bring tools and policies; the harness runs the loop)
Self-hosted multi-agent orchestration service: a lead agent delegates queued tasks to Claude Code, Codex, and other workers in isolated containers. The harness contribution is the lead-to-worker hierarchy, the task queue, shared memory, and review gates, with MCP tools and Slack and GitHub integrations; the workers own their inner loops. mcp · memory · multi-agent · sandbox · typescript
Hierarchical coding-agent harness: each autonomous loop works in its own git worktree and can delegate separable subtasks to child loops, so the execution tree grows recursively, with five coding-agent CLIs wired into per-node loops, shared SQLite run state, operator controls, and depth, cost, and time limits. python
Local-first control plane that routes one coding thread across Claude Code, Codex, Cursor, and OpenCode. The harness contribution is the layer around those agents: quota-aware rotation across isolated subscription profiles, bounded context handoffs between agents and accounts, best-of-N races, cross-family review, and typed apply gates. multi-agent · sandbox · ide · typescript
✅
slightly complex (multi-harness control plane, best-of-N, apply gates)
Multitask, multiturn RL for LLM agents; Ray-based scaling, rollout/actor workers—for teams that want to train agents, not just run them. training · python
macOS canvas that runs Claude Code, Codex CLI, Copilot CLI, and OpenCode sessions as nodes on a graph. The harness is the graph itself: each loop is typed by what ends it (a turn, a shell predicate or declared goal, a cadence, or a sub-graph), hand-off and spawn edges fire when a loop resolves, and a launchd daemon keeps sessions alive across app quits and reboots; the agent CLIs own their inner loops. memory · workflow · cli
Control plane for teams of agents on your own Kubernetes cluster, deployed with Helm. The harness contribution is the layer around runtimes such as Hermes and OpenClaw: chat front ends (Slack, Teams, Telegram, Discord, email), tool wiring to Jira, GitHub, and Confluence, a model key per agent, and an audit record of every conversation, tool call, and cost. python
✅
complex (Kubernetes control plane, multi-channel, audit, product suite)
The official reference collection of Model Context Protocol servers (filesystem, git, fetch, memory, time, and more)—the canonical, vetted toolset agents connect to, and the pattern every other MCP server is measured against. mcp · memory · typescript
MCP server that injects up-to-date, version-specific library docs into an agent's context on demand; kills the stale-training-data hallucinations that plague codegen. mcp · training · typescript
Google's official Chrome DevTools MCP server: exposes console, network, and performance-trace inspection as tool calls agents can drive directly, instead of a human clicking through the DevTools panel. mcp · browser · typescript
Playwright's official MCP server: structured browser control (navigate, click, fill, extract) via the accessibility tree rather than screenshots, so web tasks stay fast and deterministic. mcp · vision · browser · typescript