{"slug":"dark-factory","title":"dark-factory","description":"Autonomous AI development pipeline that ships code without human intervention. \nImplements the Dark Factory pattern with governance tiers (T0-T4), multi-run \nholdout validation with quorum voting, two-pass adversarial satisfaction testing, \nand the Ralph Wiggum loop for continuous backlog consumption. Each task runs in \nfresh context to prevent drift, with automatic guardrails that learn from past \nfailures. Configurable risk scoring routes low-risk changes to auto-merge while \nescalating high-risk changes for human review.","developer":null,"category":null,"reported_installs":null,"active":true,"first_seen_at":"2026-10-03T22:13:33.942Z","last_seen_at":"2026-10-08T06:00:01.054Z","last_changed_at":"2026-10-03T22:13:33.942Z","platform":"claude","sources":["Community"],"works_with":["Claude Code"],"marketplace_url":null,"repository_url":"https://github.com/everva/claude-plugins.git","web_data":{},"github_data":{},"catalog_sources":{"community":[{"name":"dark-factory","source":{"sha":"bd322d71c628bfc5a0914ac1c9354552fbc2f023","url":"https://github.com/everva/claude-plugins.git","source":"url"},"homepage":"https://github.com/everva/claude-plugins","description":"Autonomous AI development pipeline that ships code without human intervention. \nImplements the Dark Factory pattern with governance tiers (T0-T4), multi-run \nholdout validation with quorum voting, two-pass adversarial satisfaction testing, \nand the Ralph Wiggum loop for continuous backlog consumption. Each task runs in \nfresh context to prevent drift, with automatic guardrails that learn from past \nfailures. Configurable risk scoring routes low-risk changes to auto-merge while \nescalating high-risk changes for human review."}]}}