Iris MCP gives coding agents runtime evidence, not screenshots
Iris is an open-source MCP tool that lets coding agents verify their changes against a real running app — checking API responses, UI events, and console errors — rather than relying on screenshots.
Score breakdown
Iris replaces screenshot-based or assumption-based verification with runtime evidence from a live app, giving coding agents a concrete, structured verdict on whether their changes actually worked.
- 01Iris is an open-source MCP tool by syrin-labs that verifies coding agent changes against a real running app at runtime.
- 02It checks for concrete outcomes: API calls returning 200, modals opening, route changes, absence of console errors, and webhook arrivals.
- 03On React projects, Iris reports the exact `file:line` to fix when something breaks.
Iris is an open-source, TypeScript-based MCP tool published by syrin-labs that provides coding agents with runtime verification of their changes. Rather than relying on screenshots or static analysis, Iris connects to a real running app and checks that observable outcomes — such as a 200 API response, a modal opening, a route change, the absence of console errors, or a webhook arriving — actually occurred after a code change. When a failure is detected, Iris returns a structured verdict describing what broke, why, and on React projects, the specific `file:line` responsible.
The project is described as dev-only and localhost-only, and is built with TypeScript using the Model Context Protocol.
The project is described as dev-only and localhost-only, and is built with TypeScript using the Model Context Protocol. Installation is designed to be agent-driven: pasting a reference to `SKILL.md` into an AI tool causes the agent to detect whether Iris is already configured (by checking for `.iris.json`), running a setup wizard on first use and running verification on every subsequent use. The repository had 57 stars and 3 forks at the time of posting, and the README references a claim that the approach is approximately 73× cheaper than alternatives, though the source text is truncated before that detail is explained.
Key facts
- 01Iris is an open-source MCP tool by syrin-labs that verifies coding agent changes against a real running app at runtime.
- 02It checks for concrete outcomes: API calls returning 200, modals opening, route changes, absence of console errors, and webhook arrivals.
- 03On React projects, Iris reports the exact `file:line` to fix when something breaks.
- 04The tool is TypeScript-based, dev-only, and localhost-only.
- 05Installation is agent-driven: pasting a reference to `SKILL.md` causes the agent to auto-detect setup state via `.iris.json`.
- 06The repository had 57 stars and 3 forks at the time of posting.
- 07The README references a claim that the approach is ~73× cheaper than alternatives, but the source is truncated before the explanation.
Topics
Summary and scoring are generated automatically from the original article. We always link back to the publisher and never republish images or paywalled content. Last processed Jun 19, 2026 · 10:25 UTC. How this works →