Decision Review API for AI Agents
Deepam_Kapur · reddit · 2026-07-18
The author is developing an API/MCP that allows AI agents to submit "critical questions or proposed actions" to multiple independent models for review.
The API returns:
- Points of agreement and disagreement
- Unsupported assumptions
- The strongest counter-arguments
- Confidence levels and missing information
- Suggested next steps
Envisioned use cases include idea validation, software architecture reviews, second opinions, and business decision audits. Future plans include support for hosted models, BYOK (Bring Your Own Key), and local model modes. A 10-person beta is launching next week.
Related event: Concept for an MCP to Review AI Agent Decisions(2 posts)→
More from coding & agent
- Devin adds e2b sandboxes for remote agent execution — badphilosopher · 2026-07-22
- Hermes Agent Refactoring Proposal: Decoupling via Event Bus and Monorepo Slicing — Promptmethus · 2026-07-22
- ty now reads Pydantic config keywords and field metadata — charliermarsh · 2026-07-22
- Pensar Launches AI Security Agent to Autonomously Discover and Patch 0-Days — andriy_mulyar · 2026-07-22
- ty adds first-class Pydantic support, including strict and lax field handling — charliermarsh · 2026-07-22
- Google launches Gemini 3.5 Flash Cyber for CodeMender, with limited access for governments — GoogleAI · 2026-07-22