hitl

Orchestrate bounded, resumable human-in-the-loop reviews with durable state and apply-policy.

9|3|Updated Apr 9, 2026
One-click install
npx skills add https://github.com/corca-ai/charness --skill hitl
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hitl
Source: https://github.com/corca-ai/charness/tree/main/plugins/charness/skills/hitl
Command: npx skills add https://github.com/corca-ai/charness --skill hitl

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Use when automated review is not enough and deliberate human judgment needs to be inserted into a bounded review loop. Keeps review state resumable, chunked, and adapter-driven without hardcoding one host runtime.

Core Features & Use Cases

  • insert deliberate human judgment where automation is insufficient
  • keep the review bounded and resumable with durable artifacts
  • record rules and decisions so the loop can continue coherently

Quick Start

Initialize a HITL session in your repository to begin a bounded, resumable human review workflow.

Frequently Asked Questions about hitl

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add human review steps to an automated codebase workflow?

To add human review to an automated codebase workflow, initialize a bounded, resumable human-in-the-loop session. This orchestrates deliberate human judgment using a repo-owned runtime and adapter to persist durable state and decisions.

What is a bounded human-in-the-loop review workflow?

A bounded human-in-the-loop review workflow inserts deliberate human judgment into automated processes. It keeps review state chunked and resumable with durable artifacts, recording rules and decisions so the loop can continue coherently.

Do I need an adapter to manage review state in my repository?

Yes, you need an adapter and runtime artifacts to manage review state in your repository. The adapter persists state and decisions, while the runtime provides durable guardrails and an apply-policy to govern when edits are committed.

How do I resume an interrupted code review process?

You resume an interrupted code review process by relying on the durable artifacts and adapter-driven state management. The resumable human-in-the-loop runtime records previous rules and decisions so the review loop can continue coherently.

When should I use deliberate human review instead of fully automated changes?

Use deliberate human review instead of fully automated changes when automated review is not enough and human judgment is needed. It provides explicit human oversight with an apply-policy to govern exactly when automated edits are committed.