witness-run

Orchestrate witness-ratchet runs with preflight checks, worktrees, and redteam reviews.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/Elijah-J/InfoScraper --skill witness-run
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: witness-run
Source: https://github.com/Elijah-J/InfoScraper/tree/main/.agents/skills/witness-run
Command: npx skills add https://github.com/Elijah-J/InfoScraper --skill witness-run

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

The Witness Run Skill provides a structured, auditable orchestration layer to manage witness-ratchet runs, ensuring a single point of truth and disciplined progression through preflight, candidate worktrees, and red-team reviews.

Core Features & Use Cases

  • End-to-end run orchestration: coordinates preflight validation, candidate materialization, worktree management, and transition rules for witness-ratchet experiments.
  • Durable bookkeeping: centralizes state, logs, and artifacts in a single responsible cockpit, preventing drift between steps.
  • Controlled resumes and pausing: supports --resume flows and bounded pausing with explicit guardrails to manage long-running runs.

Quick Start

Initiate a witness ratchet orchestration for a given run slug with optional --resume and --max-accepted parameters.

Frequently Asked Questions about witness-run

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

FAQPage Schema
How do I orchestrate end-to-end witness-ratchet runs with strict validation?

Orchestrate witness-ratchet runs by coordinating preflight checks, candidate worktrees, and redteam reviews to ensure rigorous governance. This approach centralizes state, logs, and artifacts to prevent drift between steps. It applies deterministic task sequencing for safe experimentation.

What is the best way to manage state and artifacts during witness-ratchet experimentation?

Manage state and artifacts during witness-ratchet experimentation by centralizing bookkeeping in a single responsible cockpit. This prevents drift between steps and provides an audit-friendly record. Durable evidence capture ensures all transitions are strictly controlled.

Can I pause and resume long-running witness-ratchet workflows?

Yes, you can resume long-running witness-ratchet workflows using the --resume flow with bounded pausing. Explicit guardrails manage controlled resumes and pausing. This ensures strict validation and durable evidence capture are maintained throughout the process.

How do preflight checks and redteam reviews work in witness-ratchet orchestration?

Preflight checks and redteam reviews work in witness-ratchet orchestration by enforcing controlled transitions and strict validation. The orchestration layer coordinates preflight validation, materializes candidates in worktrees, and sequences red-team reviews for audit-friendly governance.

Do I need specific dependencies to run modular witness-ratchet orchestration?

No specific dependencies are required to run modular witness-ratchet orchestration. The system supports modular extension through reference materials, allowing you to apply deterministic task sequencing and controlled transitions directly within your existing environment.

When should I not use a structured orchestration layer for witness-ratchet runs?

You should not use structured orchestration for witness-ratchet runs when strict validation, durable evidence capture, or audit-friendly bookkeeping are not required. If your experimentation does not need controlled transitions or deterministic task sequencing, this governance layer is unnecessary.