HN Showcase
Back to browse
OpenSOP, We got tired of agents lying to us, so we built them a harness

OpenSOP, We got tired of agents lying to us, so we built them a harness

by carlosamg·Jun 3, 2026·5 points·3 comments
Visit ProjectView on HN

AI Analysis

●●SolidBold BetShip It

YAML process harness that blocks LLMs from skipping steps with receipts.

Strengths
  • •Deterministic gates prevent agents from lying about completed work
  • •Auditable receipts per step enable debugging and replay
  • •5-10x cost reduction by reserving LLM calls for actual intelligence
Weaknesses
  • •v0.1 developer preview with only 3 stars means unproven in production
  • •Agent reliability market is crowded with LangGraph, CrewAI, and others
Category
AI/ML
Target Audience

AI engineers building production agent workflows

Similar To

LangGraph · CrewAI · Temporal

Post Description

OpenSOP is an early open-source runtime/standard for executable agentic processes.

You (or your agent) define a process in YAML, and OpenSOP exposes it as a typed REST API that agents and humans can both use.

We built it because a lot of agent workflows still live in prompts, docs, or one-off scripts instead of versioned process definitions, and we wanted more control and auditability. Its under development, we are using it in production (at Coba.ai), feedback on the model, API shape, and use cases would be very useful.

We wanted to share it with the community, any feedback and comments are welcome. Especially now that Claude has released Ant https://x.com/ClaudeDevs/status/2061877343078244459 and its around the same concept.

Similar Projects