---
title: "Discussion on \"ClearCode Part 4: Autonomous Plan Execution, LLM-as-Judge, and Human-in-the-Loop Approval\" | Hashnode"
description: "Discussion on \"ClearCode Part 4: Autonomous Plan Execution, LLM-as-Judge, and Human-in-the-Loop Approval\". Where Part 3 left us\nPart 3 built four capabilities: short-term memory with session persistence, multi-step agent reasoning across tool calls, a three-tier progressive skills system, and MCP integrati"
url: "https://hashnode.com/posts/clearcode-part-4-autonomous-plan-execution-llm-as-judge-and-human-in-the-loop-approval/6a546579b9e83866c518699f"
---

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

Discussion

# [ClearCode Part 4: Autonomous Plan Execution, LLM-as-Judge, and Human-in-the-Loop Approval](https://blog.divyampatro.dev/clearcode-part-4-autonomous-plan-execution-llm-as-judge-and-human-in-the-loop-approval?utm_source=hashnode&utm_medium=feed)

Where Part 3 left us Part 3 built four capabilities: short-term memory with session persistence, multi-step agent reasoning across tool calls, a three-tier progressive skills system, and MCP integrati

Jul 13 · 9 min read · blog.divyampatro.dev  
Read the article

-   [#ai-agents](https://hashnode.com/tag/ai-agents)
-   [#build-in-public](https://hashnode.com/tag/build-in-public)
-   [#langgraph](https://hashnode.com/tag/langgraph)
-   [#python](https://hashnode.com/tag/python)
-   [#autonomous-agents](https://hashnode.com/tag/autonomous-agents)
-   [#task-planning](https://hashnode.com/tag/task-planning)

## Written by

[Divyam Patro](https://hashnode.com/@divyamp)

@divyamp · Seattle, USA

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

No fluff. Fullstack, system design, and AI — explained precisely

Fullstack Software Engineer specialising in cloud architecture, system design, and AI. I write deep dives on building production systems — the decisions, the tradeoffs, and the reasoning behind them. No fluff, no tutorials you've seen a hundred times.

## Responses 1

Markdown supported · ⌘↵ to post

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

[Jul 13](https://hashnode.com/posts/clearcode-part-4-autonomous-plan-execution-llm-as-judge-and-human-in-the-loop-approval/6a546579b9e83866c518699f/comment/6a55338dcdea897c485b615c)

Least-privilege tools per task plus an LLM-as-judge inside the executor is a strong pairing, you bound what each step can touch and score its output before it counts as done. The recovery detail, resetting IN\_PROGRESS to PENDING on restart, is what people skip until a crash mid-task wedges the whole plan. For the judge, is it scoring each task against the plan's stated intent, and can a failed score reopen the task or does it only flag for the human A/M/R loop?
