Open-source Kaktoos verifies AI-written API integrations against OpenAPI contracts via MCP
richocolate · reddit · 2026-09-11
A developer released Kaktoos, an open-source tool targeting a specific pain point of AI coding agents: agents write API integrations fast, but when the same agent also writes the tests, passing tests don't prove the integration matches the real API.
Kaktoos takes an independent verification approach: after the agent generates integration code, Kaktoos executes multi-step API workflows, checks responses against the OpenAPI contract and the real API, detects issues like missing required fields, wrong types, or unexpected status/content types, and feeds structured failures back to the agent to fix.
Key features:
- Verification exposed via MCP, so coding agents can call it directly
- The same verification runs in GitHub Actions (CI)
- The core experiment: is independent, third-party verification valuable when code is AI-generated?
The author is seeking early feedback on whether the workflow is usable, whether defining scenarios is too much overhead, and whether it catches bugs existing tests miss. Repo: KaktoosLabs/kaktoos.
More from coding & agent
- DeepSeek v4.1 Flash Tested Across 8 Coding Harnesses: Performs Best in Minimal Setups — mariofilhoml · 2026-09-11
- Sophia Yang wants a meta-harness to orchestrate Claude, Codex, and Grok agents — sophiamyang · 2026-09-11
- GPT-6 Astra talks to Muse agent to spin up a Tailscale bridge into Things 3 — SeanOliver · 2026-09-11
- Automatic1111 Returns With Workflow1111: Run 11 AI Pipelines Without 20GB of Checkpoints — Gradio · 2026-09-11
- Traces Launches Surfaces to Visualize Agent Session Activity, Token Use and Security Issues — Scobleizer · 2026-09-11
- Dev Reports DeepSeek-V4.1-Flash Breaks 100-Step Clinical Agent Workflow Old Version Ran Fine — MaziyarPanahi · 2026-09-11