8 other AI Coding tools we have sourced reviews for, all in the same niche as Agent Marketplace (Code Generation). Within that, they're ordered by aitrove popularity, not by a quality score: we haven't tested them hands-on, so we don't rank them as "best".
Agent Marketplace in one minute
Best for: Teams that use several AI coding tools and want one shared, versioned set of skills and MCP configs.
Biggest catch: Hooks support is marked experimental.
Skip it if: You use only one coding agent, which can already manage its own config.
Price: Free and open source (Apache-2.0). (checked Oct 23, 2026)
Developers who want their coding agent to prove a UI change actually works in the running app before it says 'done'.
It instruments your app: a dev-only SDK, build-config changes, edits to CLAUDE.md/AGENTS.md and prompt hooks in each agent's config
Free and open source: SDK, adapters and core are Apache-2.0; the server and CLI are FSL-1.1-ALv2 (source-available). The optional hosted Harness/dashboard has no published price. (checked Oct 21, 2026)
Developers and teams who want a habit check that they understand agent-written changes.
Your changed code goes to the reviewer provider, and there's no default cap on model calls (about 5 to 8 requests per check)
Free and open source (MIT). Reviews run on your Claude Code or Codex plan, or a separately billed OpenRouter or OpenAI-compatible API. (checked Oct 19, 2026)
ML-curious programmers interested in how programs map onto neural network weights.
A research experiment, not a production tool
Free and open source (MIT). (checked Oct 23, 2026)
Best for different needs
If you need…
Worth a look
Open-source or self-hosted
fframes— developers and coding agents who want programmatic, GPU-fast motion video in Rust. BlastLoop— developers who want a lightweight, structured build-and-review loop for agent-written code. agent-usage-guard— claude Code Pro or Max users who keep hitting their usage window mid-task.
Terminal / CLI-first
Reticle— developers who want their coding agent to prove a UI change actually works in the running app before it says 'done'.
API for developers
Plimsoll— developers building agents or MCP tools that execute model-written code and want auditable isolation and credential-free API access. Sphinx— developers and teams who want a habit check that they understand agent-written changes.
Team or enterprise features
Sphinx— developers and teams who want a habit check that they understand agent-written changes.
Groupings are generated from each tool's stated capabilities and pricing in our sourced profiles — not hands-on testing.
Best for: Developers who want their coding agent to prove a UI change actually works in the running app before it says 'done'.
Pros
Returns a verdict (yes, no or 'unknown') with the network request, state change or console line behind it and the file:line to open, instead of a screenshot
Works through MCP with Claude Code, Cursor, Codex, Windsurf, Gemini CLI and others, on web apps (React, Next.js, Vue, Svelte, Angular) and Electron/Tauri desktop apps
Saved flows replay without an AI model, and 'reticle gate' can fail a CI change that touches a saved flow without a new passing run
Biggest catch
It instruments your app: a dev-only SDK, build-config changes, edits to CLAUDE.md/AGENTS.md and prompt hooks in each agent's config
Skip it if: You need to test sites you don't own or a cross-browser matrix, where Playwright is still the better fit.
Starting price: Free and open source: SDK, adapters and core are Apache-2.0; the server and CLI are FSL-1.1-ALv2 (source-available). The optional hosted Harness/dashboard has no published price. (checked Oct 21, 2026)
Best for: Developers building agents or MCP tools that execute model-written code and want auditable isolation and credential-free API access.
Pros
Every run reports the isolation tier it actually got (process, container, kernel or vm), and a request can set a minimum tier so nothing runs on a weaker sandbox
Capability grants let agent-written code call your API through a broker that attaches a per-run credential, so the sandbox never holds your key or open network access
One self-hosted Go daemon switches backends by config: WebAssembly/QuickJS or Docker locally, gVisor, E2B or Docker Cloud microVMs in production, with clients for Vercel AI SDK, Google ADK, CrewAI, Agno, Mastra and Trigger.dev
Biggest catch
Very young project: about 25 GitHub stars and one fork when checked, built by one developer for their own projects
Skip it if: You want a hosted, zero-ops code interpreter, or you need to sandbox a whole agent and its file system rather than individual code runs.
Starting price: Free and open source (Apache-2.0); optional E2B or Docker Cloud backends bill separately per run. (checked Oct 13, 2026)
Best for: Developers and teams who want a habit check that they understand agent-written changes.
Pros
Pauses pushes to protected branches and quizzes you on what the diff does when things go wrong before letting it through
Questions come from a separate reviewer model that never saw the chat that wrote the code
Free and MIT-licensed; works with Claude Code, Codex, Cursor, OpenCode, Pi or plain Git
Biggest catch
Your changed code goes to the reviewer provider, and there's no default cap on model calls (about 5 to 8 requests per check)
Skip it if: You need an enforced review gate, or you can't send code to a model provider.
Starting price: Free and open source (MIT). Reviews run on your Claude Code or Codex plan, or a separately billed OpenRouter or OpenAI-compatible API. (checked Oct 19, 2026)
Pros, catches and prices are compiled from public sources (official pricing and docs pages, published reviews, user reports), not hands-on testing. Prices change often, so check the vendor page before you buy.