forge

Bootstrap Forge in the current project. Detects project type, configures validation commands, installs pre-commit hook, and vendors ralph-loop. Use this when setting up a new project for Forge spec runs and bead execution.

Create a subject-named specification from any evidence source using a reducer-based Forge workflow. Use this when the user wants a planning-ready spec, a clean-room reverse spec, or an evidence-first feature spec with sub-agent fanout, provenance tracking, adaptive clarification, speculative variants, and a canonical readiness contract.

Stamp a self-replicating four-bead autoresearch loop (doer/judge/arbiter/strategist) for autonomous iterative improvement of any artifact. Use after interactive intake when you have a program and want to run an autonomous improvement loop with blind scoring. Triggers on: autoresearch, research loop, autonomous loop, overnight loop, iterative improvement.

Decompose an implementation plan into br beads with dependency wiring, epic grouping, and provenance labels. Use this after spec-plan-handoff when the user accepts the beads generation prompt.

Stress-test beads for coverage, granularity, dependency correctness, and actionability using adversarial agent teams. Use this after spec-beads-generate when the beads workspace is populated.

Score a subject spec against the ontology and canonical readiness contract. Use this when you need claim-level coverage scoring, corroboration checks, critical-decision coverage, risk-weighted assumptions, contradiction penalties, and blocker-aware 80/80 gates.

Normalize all starting evidence into a subject spec run. Use this when beginning a Forge spec from code, docs, transcripts, screenshots, URLs, or a sparse user request. Creates the subject slug, frontmatter, evidence ledgers, request archetype, evidence-density classification, and canonical readiness skeleton.

Run the evidence-first Forge loop for a subject spec. Use this when you need to process evidence one unit at a time, produce unit summaries, rewrite the rolling summary, branch out to sub-agents for bounded exploration, choose adaptive clarification profiles, and emit speculative variants when blockers remain.

Render the implementation handoff for a subject spec. Use this when the canonical readiness verdict already exists and you need either a real implementation plan for eligible runs or a withheld handoff plus bounded options for blocked runs.

Reduce typed sub-agent outputs into canon and verify the subject spec before readiness promotion. Use this when you need deterministic merges, provenance review, contradiction handling, verification loops, section drafting, or final review passes across the spec.

Stress-test a subject spec for ambiguity, gaps, contradictions, and untestable claims using dynamic agent teams. Use this when the spec has passed completeness and synthesis-review gates and needs adversarial validation before readiness promotion.