# chaseai-yt/claudex-loop

Claude Code skill: four-phase plan hardening (recon, interrogate, Codex adversarial review, cross-model build & inspection) — two AI models harden your plan before a line of code exists, then swap jobs to build it. Whoever built it never grades it.

Repository: https://github.com/chaseai-yt/claudex-loop
Canonical: https://ross.abutalabs.com/products/claudex-loop
License: NOASSERTION
License Family: other
Last push: 2026-08-24T19:51:37+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 35, longevity 6
- inputs: {"age_days": 90, "days_push": 9, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1312, forks 132 (observed 2026-08-28T04:04:19.944971+00:00)

## What it is
A Claude Code skill and plugin that hardens software plans through a four-phase loop: recon, interrogation, adversarial review by OpenAI Codex, and cross-model build with inspection. Two rival AI models stress-test a plan before implementation, then swap roles so the builder never grades its own work.

## Use cases
- harden an AI-generated implementation plan before coding
- find flaws in a plan using a second AI model as adversarial reviewer
- avoid echo-chamber self-review when Claude writes and grades its own plan
- run a structured interview to lock down software requirements
- have one AI build a feature and a different AI review the diff
- validate that a plan is actually buildable before writing code

## When to choose
- you use Claude Code and want cross-model plan validation with OpenAI Codex
- AI-built plans keep containing unbuildable designs or subtle bugs
- you want a structured, human-gated workflow for AI-assisted development
- you need adversarial review where the reviewer is read-only and independent

## When to avoid
- you don't use Claude Code or OpenAI Codex
- you want a fully autonomous coding agent without human checkpoints
- you need a lightweight linter or CI-only review tool
- your license requirements demand a standard OSI-approved license

## Facets
- artifact type: plugin
- maturity: active
- function: agent-framework, prompt-engineering, code-review, developer-tools, workflow-automation
- domain: developer-tools, artificial-intelligence
- platform: cli, cross-platform
- tags: claude-code, claude-code-skill, openai-codex, adversarial-review, plan-hardening, cross-model-review, ai-coding-workflow, ai-agents, automation

## Member repositories
- chaseai-yt/claudex-loop (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:19.944971+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:48:57.352279+00:00, confidence not recorded.
  - readme: https://github.com/chaseai-yt/claudex-loop (fetched 2026-08-28T04:04:19.944971+00:00, sha 935c43ecde45)
- Data as of 2026-08-30T08:39:29.467469+00:00.