- docs/showcase/ai-org.md
- README.md
- ARCHITECTURE.md
- VISION.md
Project overview
Architecture phase: blueprint, pipeline status, artifacts, and phase settings in one work surface. 
Product requirements: versioned PRD from the purpose flow with approval status and artifact list. 
Prompt editor: system template, model settings, and agent instructions inside the run context. Verified artifacts Open risks Present Still missing Architecture / maturity Style note
Build module
AI-Org Prototype 2.0
AI-Org experiments with a delivery pipeline where agents propose and humans hold the binding gates.
DraftArchitecture review neededCharter: Concierge pilot · draftAI / AutomationCurated after review
Product screenshots
Pipeline and prompt screens ready
Vision
Prototype
Agents propose. Humans decide.
AI-Org demonstrates an agentic delivery pipeline while keeping every binding assessment human-reviewed.
The needed contributions turn a strong prototype into a reviewable demo with a narrow domain.
Why now
Without external architecture review, scope will drift. With a narrow spike, AI-Org can show where Contributors needs real review gates.
Public audit summary
Short public framing. Internal notes and reviewer comments stay in the audit case.
- Audit status
- C - In Review
- Contribution readiness
- Multi-agent architecture mentor and 5 secondary ask contexts are structured for curated contribution review.
- Last reviewed
- 2026-05-15T11:00:00.000Z
- Human gate
- AI-Org Prototype 2.0 was human-reviewed as a static beta read-model baseline; AI scorecards remain non-binding.
- Complete architecture red-team review before open contribution flow.
Structured ask list
Lead ask first, followed by secondary missions with capability, status, evidence, and value.
Lead ask01
mentor-domain-expertLead askExpert reviewIn review
Multi-agent architecture mentor
A multi-agent, graph, or orchestration architect who reduces the nine-phase plan around scope, state, eval, and human gates.
- Why now
- The prototype needs boundaries and a testable spike before more automation layers appear.
- Format
- Architecture red team with annotated pipeline, kill criteria, and spike plan.
- Not meant as
- No autonomous agent factory, no demo hype without reviewable gates.
- Owner
- Jonas Weber
- Participation
- Voluntary review / CP contribution record, no automatic payment
- Legal status
- Platform review cleared for pilot framing
People & contributions
Public roles and cleared contribution records for this project.
Expert reviewHuman gate review
Jonas Weber140 CPHuman gates, eval boundaries, and evidence requirements are being structured for the AI-Org spike.
Architecture / maturity
The build module: shows how far AI may help before Contributors sets human review and ledger gates.
Nine-phase plan, agent idea, human-in-the-loop doctrine, and strong Contributors coupling.
Narrow pilot domain, state model, eval plan, demo hosting, and UI narrative.
The nine-phase plan and multi-agent idea are described; what is needed is an experienced architecture check that narrows the pipeline.
The surface must show automation soberly: run, evidence, gate, reviewer, not science fiction.
InitiatorTim Mergenthaler
Initiator of Contributors Community and several beta projects; reviewer for governance, product, and AI boundaries.