How does M3-Anchor prevent production errors?

By forcing the identification of constraints—business, technical, and operational—before coding starts, it surfaces potential blockers and architectural gaps that usually only appear during production.

What is a 'Required Truth' (RT)?

A Required Truth is a specific condition or capability that must exist for a feature to work as intended, such as an idempotent retry mechanism or a specific error taxonomy.

Does this skill generate source code directly?

No, M3-Anchor is a planning and anchoring skill. it prepares the 'ANCHORED' state and identifies the best solution path, which then informs subsequent code generation phases.

What is backward reasoning in the context of Manifold?

Backward reasoning is a methodology that starts with the desired end goal and works in reverse to identify every condition or 'Required Truth' that must be satisfied for that goal to be achieved.

Manifold Outcome Anchoring

Name: Manifold Outcome Anchoring
Author: dhanesh

bydhanesh

•

Productivity & Workflow

Derives project requirements through backward reasoning to identify critical dependencies and hidden constraints before implementation.

Manifold M3-Anchor is a specialized Claude Code skill designed to eliminate requirement discovery during the implementation phase by enforcing a rigorous backward-reasoning workflow. By starting with a defined outcome and asking 'What must be TRUE?' for that goal to succeed, the skill systematically identifies Required Truths (RTs), maps them to existing constraints, and surfaces critical gaps. This proactive approach ensures that architectural decisions are grounded in necessity rather than assumption, providing a structured solution space with ranked options and complexity ratings before a single line of code is written.

Key Features

01Automated Required Truth (RT) identification and status tracking

02Built-in phase validation to ensure schema and architectural compliance

03Dual-output system (JSON for state tracking, Markdown for readability)

043 GitHub stars

05Backward reasoning engine for systematic requirement derivation

06Solution space generation with gap analysis and complexity ratings

Use Cases

01Translating high-level business objectives into specific technical specifications

02Evaluating multiple implementation strategies against a set of non-negotiable constraints

03Architecting complex systems where edge cases are often missed during initial planning

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add dhanesh/manifold manifold-m3-anchor

For use in Claude.ai and ChatGPT

Download Skill