# Dispatches from O'Reilly: From capabilities to responsibilities

[Stack Overflow](https://yomu.fyi/company/stack-overflow) · Artur Huk · Jun 19, 2026

**Type:** Problem & solution

## Summary

High-stakes AI agents capable of mutating external state often face governance failures when relying on system prompts or manual Human-in-the-Loop approval queues that quickly degrade into alert fatigue. The Responsibility-Oriented Agent architecture addresses this operational bottleneck by shifting system design from open-ended capability framing to deterministic, contract-enforced responsibilities. Under this model, underlying orchestration frameworks like LangChain operate in User Space with their side-effecting tools removed, isolating the agent to epistemic reasoning. The agent expresses its intended action exclusively by emitting a structured policy proposal to a privileged Kernel Space runtime. The runtime deterministically evaluates the proposal against versioned YAML contracts registered in an agent registry, ensuring that only genuine policy exceptions are escalated to human supervisors.

## Context

High-stakes agentic AI systems that mutate external state, such as executing trades or modifying infrastructure, become operational bottlenecks when managed via Human-in-the-Loop review queues. Reviewers experience alert fatigue, while relying on prompt-based guardrails fails to deterministically constrain authority.

## Approach / What changed

The Responsibility-Oriented Agent pattern establishes a deterministic Kernel Space that evaluates actions against machine-readable Responsibility Contracts. Agents are stripped of direct execution tools and restricted to emitting structured policy proposals, escalating only contract violations to humans.

## Takeaways

- Human-in-the-Loop approval queues scale poorly in high-frequency production systems, degrading into alert fatigue where operators approve payloads without thorough verification.
- The Responsibility-Oriented Agent pattern decouples reasoning from execution, requiring agents to emit structured policy proposals rather than calling state-mutating APIs directly.
- Authority is bounded by versioned, machine-readable contracts evaluated by a deterministic kernel, replacing binary manual queues with governance by exception.

**Tags:** [Architecture](https://yomu.fyi/topic/architecture), [LLMs](https://yomu.fyi/topic/llm), [Reliability](https://yomu.fyi/topic/reliability)

[Read original post](https://stackoverflow.blog/2026/06/19/dispatches-from-o-reilly-from-capabilities-to-responsibilities)