Is Smoke Monkey an AI model?

No. Smoke Monkey is not an LLM. It is an embeddable, framework-agnostic runtime that turns LLMs (NVIDIA, Gemini, OpenAI, Claude, Ollama) into autonomous engineering agents capable of multi-step planning, calling tools, editing files, and running test loops with zero runtime dependencies.

Do I need a database (Postgres, SQLite, or Redis)?

No database is required. State is persisted through lean session files on disk, or you can plug in your own custom storage adapter. You can deploy it as a CLI, a desktop app, or inside a serverless function without provisioning a database.

How does human-in-the-loop permission gating work?

Smoke Monkey comes with three permission policies: allow-all, deny-all, and ask-default (or a custom function resolver). When a sensitive tool like edit_file or run_command is called under "ask", the runtime emits a permission.required event, pausing execution until your code or UI calls agent.resolvePermission(id, "allow" | "deny").

How does the Model Context Protocol (MCP) server work?

Smoke Monkey ships an official stdio MCP server in smoke-monkey-harness-mcp. Point Claude Code, Cursor, Codex, or opencode at it using npx -y smoke-monkey-harness-mcp to let other AI tools plan, scaffold, verify, and run agents using Smoke Monkey automatically.

What are SKILL.md skills and how do they prevent context bloat?

Skills are directories containing a SKILL.md workflow file. The system prompt only includes a one-line catalog of available skills. When a task requires specialized expertise (e.g. frontend-ui-engineering, test-driven-development, or security-hardening), the agent calls use_skill to load instructions just-in-time, saving thousands of tokens.

What is an AI Agent Harness?

An AI agent harness is the execution environment that bridges raw Large Language Models (LLMs) to real-world software engineering capabilities. While an LLM generates tokens, an AI harness manages deterministic phase loops (explore, plan, edit, verify, recover), tool execution sandboxing, token budget compaction, and safety gates needed for autonomous problem-solving.

What makes Smoke Monkey an Agentic AI runtime rather than a simple wrapper?

Unlike basic prompt-response chains, Smoke Monkey implements true agentic AI architecture: a deterministic 6-phase state machine with automatic failure recovery, runaway loop guards, 24 built-in developer tools, dynamic subcontext isolation, and interactive human-in-the-loop permission pauses.

Is Smoke Monkey Harness 100% Open Source AI?

Yes. Smoke Monkey Harness is 100% open source under the permissive MIT license with zero runtime dependencies. There is no telemetry lock-in, no mandatory cloud backend, and no database requirement—giving engineering teams complete sovereignty over their agent code.

What Agent APIs does Smoke Monkey Harness provide?

Smoke Monkey provides composable TypeScript Agent APIs: createAgent() for initialization, agent.run() for multi-turn execution, agent.on() for fine-grained lifecycle and streaming events, agent.resolvePermission() for interactive human authorization, subcontext switching APIs, and stdio Model Context Protocol (MCP) server endpoints.

Can I run completely offline without API keys?

Yes! Smoke Monkey natively supports Ollama out of the box. Set provider: "ollama" and model: "qwen3:8b" with workspacePath, and the agent will run locally on your machine with zero external network requests.

Smoke Monkey vs Cognition Devin

Cloud Autonomous Software EngineerUpdated: October 2026

Open Source Devin Alternative: Build Autonomous Software Engineers in TypeScript

Devin popularized the autonomous AI software engineer, but charges steep monthly subscriptions and keeps its execution architecture closed. Smoke Monkey Harness provides the open-source TypeScript runtime with deterministic test loops, git management, and terminal execution.

Comparative Benchmark: Smoke Monkey Harness TypeScript vs Cognition Devin

|Verified for Node.js 18+ & Bun|100% MIT Open Source

The Executive Verdict (Quick Answer)

Why choose Smoke Monkey over Cognition Devin? Build your own autonomous software engineer without high monthly subscription fees or third-party cloud lock-in using Smoke Monkey Harness.

Why Developers Switch from Cognition Devin to Smoke Monkey

100% Free & Open Source: No $500/month seat costs. Pay only for raw LLM API tokens or run free with Ollama.

Self-Healing Test Loops: Explores, plans, edits, and runs your real test suite (`npm test`, `pytest`) until passing.

Sovereign Infrastructure: Run on your own servers or developer laptops without uploading IP to cloud sandboxes.

Extensible via MCP & Skills: Load specialized domain skills dynamically via `SKILL.md`.

Detailed Feature-by-Feature Matrix

Direct side-by-side comparison of core runtime capabilities and architectural trade-offs.

Capability	Smoke Monkey Harness	Cognition Devin	Key Architectural Difference
Monthly Cost	✅ $0 (100% Free Open Source)	❌ $500+/month enterprise seat pricing	Significant cost reduction at any engineering scale.
Self-Healing Test Cycles	✅ 6-Phase loop with automated recovery on test failure	✅ Cloud VM test execution	Smoke Monkey runs directly on your local system or container.

Code Implementation Comparison

Autonomous Test-Driven Bug Fixing Loop

Smoke Monkey (TypeScript)Zero Dependencies

autonomous-engineer.tstypescript

1import { createAgent } from 'smoke-monkey-harness';
2
3const engineer = createAgent({
4  provider: 'anthropic',
5  model: 'claude-3-7-sonnet',
6  workspacePath: process.cwd(),
7});
8
9// Cycles through edit -> test -> recover until verified
10const result = await engineer.run(
11  'Fix the failing JWT token expiration test in tests/auth.spec.ts'
12);
13console.log('Result:', result.status, 'in phase:', result.phase);

Cognition Devin$500+/mo Cloud Subscription

devin-cloud-prompttext

1// Devin executes inside Cognition's cloud VM
2// Proprietary, high waitlist latency, non-customizable runtime internals

Architecture Note: Smoke Monkey brings autonomous engineering agent loops to your personal workstation and private CI/CD pipelines.

The Verify & Recover Phase: Autonomous Self-Healing

When Smoke Monkey enters the verify phase, it runs tests or linters. If the command fails, it enters recover mode, extracting error traces, diagnosing root causes, and updating the plan before re-attempting edits.

Frequently Asked Questions

Questions Developers Ask About Cognition Devin Alternatives

Q:Can Smoke Monkey deploy code to GitHub like Devin?

Yes. Smoke Monkey includes built-in git tools: `git_status`, `git_diff`, `git_log`, and `git_commit`, allowing full VCS automation.

Switch to Smoke Monkey Harness Today

Build autonomous coding agents with zero runtime dependencies, deterministic 6-phase loops, and Model Context Protocol (MCP) in pure TypeScript.

npm install smoke-monkey-harness

Open Source Devin Alternative: Build Autonomous Software Engineers in TypeScript

Why Developers Switch from Cognition Devin to Smoke Monkey

Detailed Feature-by-Feature Matrix

Code Implementation Comparison

The Verify & Recover Phase: Autonomous Self-Healing

Questions Developers Ask About Cognition Devin Alternatives

Q:Can Smoke Monkey deploy code to GitHub like Devin?

Related Solutions & Topics

Automated Test Verification Loops: Self-Correcting Code Generation with AI

How to Prevent Infinite LLM Agent Loops: Runaway Guards & Self-Healing Phases

Free & Open Source AI Agent Framework: 100% MIT Licensed TypeScript Harness

Switch to Smoke Monkey Harness Today