human-in-the-loop

human-in-the-loop is a skill for Claude Code, Codex from hajekim/agentic-design-patterns-extension. It costs 412 tokens per session (3,623 once invoked), scanned A, a copy of human-in-the-loop, MIT.

A workflow pattern that pauses an AI agent at important checkpoints so a person can approve an action, clarify an instruction, or review a decision before work continues. HITL means human-in-the-loop.

In plain words
What is it for?
Adding approval steps to email sending, data deletion, financial or high-stakes decisions, unclear requests, and actions involving people outside the system.
Why use it?
It keeps irreversible, risky, ambiguous, or externally affecting actions from happening without human judgment.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Adding approval steps to email sending, data deletion, financial or high-stakes decisions, unclear requests, and actions involving people outside the system.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/hajekim/agentic-design-patterns-extension/human-in-the-loop
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add hajekim/agentic-design-patterns-extension --skill human-in-the-loop
Clone the repo
git clone --depth 1 https://github.com/hajekim/agentic-design-patterns-extension

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for human-in-the-loop

README.md
[![agentmods](https://agentmods.dev/badge/skills/hajekim/agentic-design-patterns-extension/human-in-the-loop/github.svg)](https://agentmods.dev/skills/hajekim/agentic-design-patterns-extension/human-in-the-loop)
Your own site
<a href="https://agentmods.dev/skills/hajekim/agentic-design-patterns-extension/human-in-the-loop"><img src="https://agentmods.dev/badge/skills/hajekim/agentic-design-patterns-extension/human-in-the-loop/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for human-in-the-loop

Your own site · 80×15
<a href="https://agentmods.dev/skills/hajekim/agentic-design-patterns-extension/human-in-the-loop"><img src="https://agentmods.dev/badge/skills/hajekim/agentic-design-patterns-extension/human-in-the-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 412 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,623 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin 100% copy Near-identical to another mod in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00412 $0.03623
Opus 5 $0.00206 $0.01811
Sonnet 5 $0.00082 $0.00725
Haiku 4.5 $0.00041 $0.00362

Measured 8d ago against content hash 012781ea2d3f, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

human-in-the-loop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

Origin

This is a copy

100% identical to human-in-the-loop — 3 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

skills/human-in-the-loop/SKILL.md · 393 lines

How it starts

The opening of the file, as written. The whole thing — 393 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Human-in-the-Loop (HITL) Pattern

Overview

The Human-in-the-Loop (HITL) Pattern integrates human judgment, oversight, and approval into agent workflows at critical decision points. Rather than running fully autonomously, HITL agents pause at defined checkpoints — to confirm high-stakes actions, request clarification, seek approval, or collect feedback — before proceeding.

Core Principle: Autonomy is earned, not assumed — build human oversight into the design, not as an afterthought.

When This Skill Applies

Activate this pattern when:

  • Agent actions are irreversible (sending emails, deleting data, financial transactions)
  • High-risk decisions require human judgment (medical, legal, financial advice)
  • Ambiguous instructions need human clarification before proceeding
  • Regulatory or compliance requirements mandate human approval
  • Agent confidence is below acceptable threshold for autonomous action
  • Actions affect third parties who have not consented to fully automated interaction
  • The agent encounters situations outside its training distribution

Rule of thumb: If you'd want a human to review the action before it happens, build in a HITL checkpoint.

HITL Spectrum

Fully Autonomous ←————————————————————→ Fully Manual
       |           |           |           |
   Auto-run    Approve     Review      Step-by-step
  (no review)  high-risk   all steps    approval

Choose the appropriate level based on:

  • Risk: Higher risk → more human involvement
  • Trust: Proven, well-tested agents can operate more autonomously
  • Context: Production vs. development, internal vs. external-facing
  • Reversibility: Irreversible actions need pre-approval

DEFINE → PLAN → ACTION Workflow

DEFINE

Map the human oversight requirements:

  1. Which actions are irreversible or high-risk?
  2. Which decision points require human judgment (not just execution)?
  3. What information does the human need to make an informed approval?
  4. What are the response time requirements? (Synchronous vs. async approval)
  5. What happens if the human is unavailable?

Read the full file on GitHub · 393 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 8d ago First seen · 393 lines · 412 tokens per session scan A 012781ea2d3f

Subscribe to this mod's changes

human-in-the-loop is a skill published in the GitHub repository hajekim/agentic-design-patterns-extension (1 stars, last pushed 5mo ago), licensed MIT. It adds 412 tokens to every session and 3,623 once invoked, about $0.0021 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to human-in-the-loop, differing in 3 lines, and is treated as a copy.