AI that doesn't guess.
For the work that matters.
Every run is checked. Every output is verified. Every decision is traceable. The Forge isn't a prompt pack — it's a governed tool with contradiction detection, honest uncertainty handling, failure blocking, and a proof receipt for every run.
Live Governance
Watch 8 agents evaluate your task through gates, witnesses, and seals — in real time
Failure Blocking
Bad input gets rejected. Contradictions get caught. Fabrication gets stopped.
Honest Unknowns
Flags what it doesn’t know instead of guessing. Says “I need more information” when it should.
Proof Receipt
Every run produces an exportable governed execution receipt — PDF, Markdown, or JSON
Governance Domains▶
Promotion Gate
Should This Be Sent?
Every piece of AI-generated content goes through a promotion gate before it can become action. The system evaluates authority, claims, audience, and risk independently.
Assumption Crusher
What Did AI Just Make Up?
When AI doesn't have enough information, it invents data and presents it with full confidence. This domain identifies every fabricated data point.
Contradiction Hunter
This Should Not Add Up
When input data contradicts itself, raw AI picks one interpretation and proceeds. This domain detects logical inconsistencies and blocks output until they're resolved.
Signal Compass
How Far Did the AI Drift?
Measures the invisible deviation between your stable reference and every AI output — showing magnitude and direction of drift in real time.
Workflow Integrity
Did the AI Skip a Step?
Enforces proper sequence when order matters — blocks actions that depend on incomplete prerequisites.
What Can You Forge?
The Forge includes 84+ pre-built governed scenarios across 8 categories. Each one shows you exactly what raw AI would produce — and exactly what governance catches before the output reaches you. Click any category to see example prompts and what governance does with them.
Sales & Lead Management
12Qualify leads, write proposals, follow up on deals — with governance that catches fabricated data, premature commitments, and contradictory inputs before they reach your prospect.
Client Communications
10Emails, announcements, status updates, and any outbound message where saying the wrong thing has real consequences. Governance checks authority, accuracy, and audience before anything is marked send-ready.
Financial & Compliance
14Budgets, reports, regulatory summaries, and financial communications. Every number is checked for source, every claim is checked for authority, and nothing goes out without the right approvals.
Content & Marketing
8Blog posts, social media, product descriptions, and marketing copy. Governance catches brand voice drift, fabricated claims, and the slow degradation of accuracy across repurposed content.
HR & Personnel
8Hiring, terminations, policy announcements, and employee communications. These are high-stakes outputs where a wrong word can create legal exposure. Governance blocks unauthorized actions and enforces required review steps.
Security & Access Control
12Risk assessments, access grants, incident response, and security reviews. The highest-stakes domain — where governance prevents AI from downplaying threats, granting unauthorized access, or skipping required verification steps.
Customer Support
8Ticket responses, escalation decisions, and customer-facing answers. Governance constrains AI to your knowledge base, scores confidence on every response, and triggers human handoff when it should.
Operations & Workflow
12Process design, automation reviews, vendor management, and operational decisions. Governance catches when AI skips required steps, contradicts your existing processes, or fabricates efficiency metrics.
Choose Your Tier▶
Month-to-month. Cancel anytime. Start free, scale when ready.
Anyone curious about governed AI output
- 3 governed runs per day
- All 5 governance domains in preview mode
- Raw vs Governed side-by-side comparison
- Basic receipt with governance signals
- Generic governance rules (not domain-tuned)
- —Run history
- —Re-Forge with memory
- —Full proof packs
- —Custom constitutions
Individual operators, freelancers, and small teams using AI daily
- Unlimited runs across all domains
- Domain-specific constitutional rules
- Full interactive Forge view (agents, gates, resonance, lineage)
- Saved run history
- Re-Forge with memory (carries context forward)
- Priority queue on re-forge
- Full proof packs (PDF / MD / JSON)
- 3 launch domains: Sales Lead Qualifier, Content Repurposer, Client Support Responder
- —API access
- —Agent kits
- —Custom constitutions
Cancel anytime
Teams embedding governance into their own workflows and products
- Everything in Pro
- API access — run Forge programmatically
- Custom constitution tuning (your own rules per domain)
- Governed Agent Kits — Docker / one-click bundles
- Prompt Forge Kits — 8-AI recursive decomposition
- Exportable proof packs for compliance and audit
- Webhook integrations
- All current + future domains unlocked
- —Control Plane dashboard
- —Flight recorder
- —Assembly registry
Cancel anytime
Founders, advanced teams, and multi-agent environments
- Everything in Engine
- Full Control Plane dashboard
- Org-wide governance architecture
- Output verification systems (pre-production gate)
- Routing and escalation logic
- Identity and permission boundaries
- Full decision lineage and audit trail
- Approval workflows for high-stakes actions
- Flight recorder (68 event types)
- Assembly registry — register your own AI assemblies
- Drift monitoring across all governed outputs
- Architecture review session (1x onboarding)
Cancel anytime
Feature Comparison
| Feature | Free | Pro | Engine | Control Plane |
|---|---|---|---|---|
| Governed runs | 3 / day | Unlimited | Unlimited | Unlimited |
| Governance domains | Preview | 3 launch | All + future | All + custom |
| Run history | — | ✓ | ✓ | ✓ |
| Re-Forge with memory | — | ✓ | Priority | Priority |
| Proof packs (PDF/JSON) | — | ✓ | ✓ | ✓ |
| API access | — | — | ✓ | ✓ |
| Agent kits | — | — | ✓ | ✓ |
| Custom constitutions | — | — | ✓ | ✓ |
| Webhooks | — | — | ✓ | ✓ |
| Control Plane dashboard | — | — | — | ✓ |
| Flight recorder / audit | — | — | — | ✓ |
| Assembly registry | — | — | — | ✓ |
| Approval workflows | — | — | — | ✓ |
Launch Domains
Available on Pro and above. Each domain is a governed AI tool purpose-built for a specific class of work.
- BANT scoring with evidence requirements
- Contradiction detection across lead data
- Disqualifier logic (not just qualification)
- Honest UNKNOWN handling — flags gaps instead of guessing
- 6-platform fact-checking
- Source-to-output traceability
- Platform-specific constraints (char limits, tone, format)
- Brand voice drift detection
- KB-only constraint — only answers from your knowledge base
- Urgency classification
- Confidence scoring on every response
- Human handoff triggers when confidence is low
What's Actually Included
Every feature listed above does something specific. Click any feature below to see what it is, why it matters, and how it works in practice.
Pro Features
$19/moEngine Features
$49/mo — everything in Pro, plus:Control Plane Features
$99/mo — everything in Engine, plus:Governance Prompt Kits
Subscriptions give you the Forge — the tool. Kits give you the governed scenarios — the library. Each kit is a collection of real AI failure modes your team will encounter, with the specific governance checks that catch them. One-time purchase. You own the templates. Every one validated through DDT-059 with auditable provenance.
AI Governance Starter Kit
- One governed scenario from each of the 9 domains below — see what governance catches in your area before committing
- Score any AI output against a 100-point rubric — know exactly where it passes and where it breaks
- Get a sample proof bundle showing what auditable AI evidence looks like
- 30-day Forge access — run governance on your own prompts, then go deep with the domain kit that matters most
Financial Governance
- Stops budget approvals when the numbers contradict the headcount plan or strategy memo
- Catches revenue claims that don’t match the attribution data behind them
- Flags compliance summaries that conflate regulatory frameworks or miss jurisdiction requirements
- Blocks financial communications until sign-off thresholds are met
Sales Governance
- Catches when AI invents revenue figures, job titles, or funding stage to inflate a lead score
- Blocks proposals that commit to pricing or timelines without current rate card verification
- Flags competitive claims that have no source, no date, or no evidence behind them
- Stops renewal proposals from ignoring usage declines or burying adverse term changes
Security Governance
- Catches vendor assessments that check the box without verifying scope, coverage, or recency
- Blocks production access grants that skip approval chains or lack time-limited scope
- Flags self-assessed security postures being presented as independently audited
- Stops incident summaries from understating severity or fabricating timeline details
Operations Governance
- Catches deployment plans that skip staging, monitoring, or rollback — before they hit production
- Flags postmortems that blame the symptom without investigating the process failure
- Blocks capacity plans where cost projections contradict the growth targets they’re built on
- Stops vendor renewals from auto-approving when adverse terms changed since last cycle
Client Communications
- Catches client emails that promise delivery dates, timelines, or scope without internal confirmation
- Blocks outbound communications where the tone drifted from factual to promotional without you noticing
- Flags client-facing documents missing required disclaimers or containing unauthorized commitments
- Stops status updates from saying "on track" when the data you provided says otherwise
Content Governance
- Catches when repurposed content introduces claims that weren’t in the original source
- Flags statistics, benchmarks, and customer quotes that have no verifiable source behind them
- Blocks press releases and case studies with superlatives that expose you to competitor or legal challenge
- Stops platform-adapted copy from drifting into claims the original content never made
HR Governance
- Blocks PIPs and termination letters that skip prior documented feedback or HR review
- Catches job postings where the requirements contradict the seniority level or each other
- Flags compensation justifications built on unsourced benchmarks or stale market data
- Stops employment actions from creating legal exposure by skipping required process steps
Support Governance
- Catches support responses that promise unreleased features, specific timelines, or roadmap items
- Blocks instant resolutions that skip verification — the fastest answer isn’t always the right one
- Flags when AI fabricates diagnostic data or invents error counts to sound authoritative
- Stops pricing corrections from creating unauthorized precedents or contractual exposure
Engineering Governance
- Blocks database migrations and schema changes that start executing before scope is confirmed
- Catches RFCs and architecture proposals built on unverified assumptions about the current system
- Flags code reviews that report "no issues" without checking specs, edge cases, or test coverage
- Stops deployment checklists from skipping staging, monitoring, or rollback prerequisites
Every kit includes governed templates with raw vs governed comparisons, specific governance actions, and auditable evidence.·Kit integrity verified via SHA-256 hash.·Direct from QenSAi — validated through DDT-059 with full provenance chain.
Forge Codex
Forge is the engine. Codex is the operator's manual. 20 rules earned through real pain points, 8 canonical procedures refined through real failure, and a coordinate system computed from real code. Not theory — extracted from a working 9-project, 3-assembly governance system.
Codex Reference
The governance playbook. Read it. Adapt it. Build your own discipline.
- 20 NTF rules — each earned through a real production pain point
- 8 canonical procedures — staged tasks, SST authority, session reports, path registry, and more
- 2VP coordinate system documentation — how every entity gets a governance position
- Session report format + Obsidian tag conventions
- File Genesis Protocol — canonical headers, Script IDs, scanner spec
- Methodology guide — how to identify and canonize your own NTFs from your own pain points
Instant access after purchase
Codex Builder
Everything in Reference plus the tooling to run governance on your own codebase.
- Everything in Codex Reference
- Measure audit scanner — scan your codebase, produce governance measure inventory
- 2VP manifest pipeline — compute coordinates from S1-S10 slots for every entity
- register-from-manifest automation — idempotent upsert to your governance database
- Staged task system — manifest-driven task lifecycle (no raw SQL)
- Entity type rollup scripts — classify by PAGE / API / COMPONENT / HOOK / LIB
- 9,236+ measure reference dataset with 2VP coordinates
Instant access after purchase
Codex + Forge Suite
Codex Builder + Complete Governance Suite. The engine AND the operator's manual.
- Everything in Codex Builder
- Complete Governance Suite — all 9 domain kits + evaluation battery + adversarial testing ($1,997 value)
- 114 governed prompt templates across all domains
- 100-point scoring rubric per output
- Proof bundle generation for every governed run
- Bundle saves $497 vs buying Codex Builder + Complete Suite separately
Instant access after purchase
Codex Subscription
Your governance framework evolves with ours. New rules. New procedures. Rolling access.
- Full Codex Reference + Builder content — always current
- New NTF rules as they're canonized from production sessions
- New canonical procedures as they land
- Updated measure reference dataset as coverage expands
- Priority access to new governance tooling
- Cancel anytime — keep what you've built
Cancel anytime
Governance in Action
Real examples from our template library. Every prompt has a lineage, a score, and a seal. These aren't hypothetical — they're from governance runs with specific findings, not generic warnings.
Budget request with hidden contradictions
A department submits a $185K budget request. The numbers add up. The AI recommends approval.
"The increase is justified by the planned product launch. The request aligns with growth targets and I recommend approval. The marketing team has consistently delivered positive ROI."
Governance found 3 contradictions the AI missed: contractor budget conflicts with new-hire plan, events budget contradicts the team's own strategy memo, and "consistent positive ROI" hides a 0.8x return on the largest line item.
Lead score built on invented data
A webinar registrant needs qualification. The AI produces a confident 82/100 lead score.
"Lead Score: 82/100 — HOT. Budget: Mid-market company (~$5M ARR). Timeline: Downloaded pricing page before registering. Company Fit: Series B. Recommendation: Fast-track to SDR."
Governance found 3 of 5 scoring dimensions were fabricated — the $5M ARR, pricing page visit, and Series B status were not in the registration data. Lead score changed from 82/100 to INCOMPLETE.
PIP that skips due process
A manager asks AI to draft a performance improvement plan. The AI produces a polished, professional document.
"PERFORMANCE IMPROVEMENT PLAN — Performance Deficiencies: Failed to meet sprint commitments, recurring quality issues, missed meetings. Consequences: Failure to meet expectations within 30 days will result in termination."
Document BLOCKED. No evidence of prior documented feedback. HR has not reviewed. Deficiencies are subjective, not specific. Improvement criteria depend on someone else's opinion. 30-day timeline may violate company policy.
To see live governance on your own prompts, try the Forge — submit any request and watch governance evaluate it in real time.
Open the ForgeEvery template validated through DDT-059 (schema · measures · glossary · severity · quality gate).·114+ governed templates across 9 domains.·Governance-backed — not just text.
Ready to Govern Your AI?
Start free. See how governed AI works in seconds. Scale when you're ready.