02. PRODUCTS

Products & open source

Every system in the series turned into something you can install, clone, or subscribe to.

ValidationForge

Functional validation for agents — evidence over mocks.

Ship functional evidence, not passing tests. Every completion claim produces screenshots, API responses, and build logs that a skeptical r…

642 test files blocked7 validation phases52 skills shippedopen ↗

Multi-Agent Consensus

Three independent agents must vote PASS unanimously — no 2/3 approvals.

Three independent agents vote on every phase gate. All three must return PASS — no 2/3 majority, no shortcuts, no skipped reviews.

929 agent spawns studied3 reviewers per gate100% unanimous requiredopen ↗

Three planning variants — linear, tournament, convergence — from rough idea to executable plan.

Three planning variants — Cast (linear), Alloy (tournament), Temper (convergence) — that turn a rough idea into a deployable implementatio…

3 planning variants98 plan score achieved17m Temper wall-clockopen ↗

The gate between 'I did the work' and 'the work is done' — 10-phase forge pipeline, 3 independent reviewers, 3 oracle auditors, zero override flags.

Crucible refuses to let the session end until 19 commands execute cleanly, 10 agents produce real-system evidence, 3 reviewers unanimously…

19 commands10 agents11 skillsopen ↗

A sharp-eyed visual-audit suite for live web UIs.

Lynx audits live web UIs through real browser sessions — two coupled skills, one verdict-writer agent, 14/14 detection accuracy across two…

14/14 detection2 skills4 commandsopen ↗

Find drift between what you asked, what you planned, what was claimed, and what shipped.

A Claude Code plugin for finding drift between what you asked, what was planned, what was claimed, and what actually shipped — sourced fro…

4 skills8 commands2 hooksopen ↗

Distill your Claude Code session history into reusable Workflow scripts.

A 3-tier session miner that turns Claude Code JSONL history into Subagents, Skills, and Workflows — classified by rubric, not by guess.

1,222 sessions mined44 clusters surfaced4 specimens polishedopen ↗

A Claude Code plugin that turns multi-stage agentic work into provable steps.

Multi-stage agentic work as a sequence of provable steps. Iron Rule: real-system evidence on disk, refuses verdicts without proof. 36 skil…

36 skills22 commands10/10 doctor checksopen →