checklist

An implementation checklist for adding HITL, or human-in-the-loop, support to an autonomous agent. It covers HTTP requests, polling, message delivery, and communication through services such as Telegram, Slack, WhatsApp, or a command line.

In plain words
What is it for?
Planning and implementing an agent that pauses for human input, checks for replies, and delivers messages across the listed channels.
Why use it?
It breaks a complex agent feature into concrete steps, so human approval or input can be added without guessing how the parts fit together.

Agent

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.

agentmods
npx agentmods add agents/rotorstar/hitl-protocol/checklist
Clone the repo
git clone --depth 1 https://github.com/rotorstar/hitl-protocol
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 4,327 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found 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 $0.00000 $0.04327
Opus 5 $0.00000 $0.02164
Sonnet 5 $0.00000 $0.00865
Haiku 4.5 $0.00000 $0.00433

Measured 2d ago against content hash 29a01e180d34, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

checklist scanned grade A with 1 finding 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 2d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

from urllib.parse import urlparse
agents/checklist.md · 445 lines

How it starts

The opening of the file, as written. The whole thing — 445 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Implementation Checklist

This document provides a step-by-step guide for implementing HITL Protocol support in an autonomous agent.

Architecture Overview

┌─────────────────────────────────────────────────────────┐
│                      Your Agent                         │
│                                                         │
│  ┌─────────────┐  ┌──────────────┐  ┌───────────────┐  │
│  │   HTTP       │  │   HITL       │  │   Message     │  │
│  │   Client     │──│   Handler    │──│   Delivery    │  │
│  │             │  │              │  │               │  │
│  └─────────────┘  └──────┬───────┘  └───────────────┘  │
│                          │                              │
│                   ┌──────┴───────┐                      │
│                   │   Poll       │                      │
│                   │   Manager    │                      │
│                   └──────────────┘                      │
└─────────────────────────────────────────────────────────┘
         │                  │                    │
    API calls          Poll/SSE            Telegram/Slack
         │                  │              WhatsApp/CLI
         ▼                  ▼                    ▼
    ┌─────────┐      ┌─────────┐          ┌─────────┐
    │ Service │      │ Service │          │  Human  │
    │  API    │      │  Poll   │          │         │
    └─────────┘      └─────────┘          └─────────┘

Decision Tree

Agent receives HTTP response from service
│
├── Status 200
│   └── Use result directly, continue workflow
│
├── Status 202 + "hitl" in body
│   ├── Extract hitl.review_url and hitl.prompt
│   ├── Send URL + prompt to human via messaging channel
│   │
│   ├── Transport: choose one
│   │   ├── Polling (default): GET hitl.poll_url every 30s-5min
│   │   ├── SSE: connect to hitl.events_url, listen for events
│   │   └── Callback: register hitl_callback_url, wait for POST
│   │
│   ├── On status "pending"
│   │   └── Continue waiting
│   │
│   ├── On status "opened"
│   │   └── (Optional) Inform human "Review page opened"
│   │
│   ├── On status "in_progress"
│   │   └── (Optional) Inform human "Reviewer is working on it"
│   │
│   ├── On status "completed"
│   │   ├── Extract result.action and result.data
│   │   ├── Check for next_case_id (multi-round)
│   │   └── Continue workflow with structured result
│   │
│   ├── On status "expired"
│   │   ├── Read default_action from hitl object
│   │   ├── Execute default_action (skip/approve/reject/abort)
│   │   └── Inform human "Review expired, using default: {action}"
│   │
│   └── On status "cancelled"
│       ├── Read reason
│       └── Inform human "Review cancelled: {reason}"
│
├── Status 4xx
│   └── Handle client error (bad request, auth, etc.)
│
└── Status 5xx
    └── Handle server error (retry with backoff)

Read the full file on GitHub · 445 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. 2d ago First seen · 445 lines · 0 tokens per session scan A 29a01e180d34

Subscribe to this mod's changes

checklist is an agent published in the GitHub repository rotorstar/hitl-protocol (11 stars, last pushed 1mo ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 4,327 tokens. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.