Roadmap
FractalMind AI has moved from protocol validation into operating-system operationalization.
Current Track: FractalMind OS v1.0
The active work now centers on making the heartbeat-driven operating loop reliable in practice:
| Focus | Why it matters |
|---|---|
| Heartbeat control loop | Keep discovery, governance, dispatch, and follow-through running without heroics |
| Structured memory | Make state durable across sessions and agents |
| Candidate OKR governance | Turn raw signals into reviewable, measurable work |
| Observability & evidence | Require logs, QA evidence, sync state, and outcome records |
| Distribution surfaces | Prepare the system to run beyond one workspace or machine |
Phase 1: Validation — Complete
Goal: prove the stack can exist end-to-end.
| Milestone | Status |
|---|---|
fractalmind-protocol live on SUI Testnet | Complete |
| First organization instance on-chain | Complete |
| Core management skills working together | Complete |
| Public docs + GitHub organization surfaces online | Complete |
Phase 2: Operating System — Current
Goal: make FractalMind usable as a governed AI operating system, not just a protocol demo.
| Milestone | Description |
|---|---|
| Heartbeat OS loop | signal -> memory -> candidate OKR -> governance -> execution -> outcome -> evolution runs as the default operating model |
| Shared strategic memory | Durable state, postmortems, and runtime memory stay coherent across sessions |
| Governed autonomy | Low-risk work can move automatically while high-risk actions stop at clear human boundaries |
| Reviewable delivery | QA evidence, reviewable change sets, and sync state become standard, not optional |
| Public narrative alignment | Website, GitHub, and docs describe the real OS-first system instead of an older protocol-only framing |
Phase 3: Distribution
Goal: let the OS run across machines and environments.
| Milestone | Description |
|---|---|
| fractalmind-envd MVP | Distributed runtime for remote agent execution |
| Gateway / bridge services | Stronger handoff between trust surfaces and off-chain execution |
| Deployment automation | One-command bootstrap of the core operating stack |
| Remote observability | Health, routing, and state remain inspectable outside one machine |
Phase 4: Federation
Goal: multiple FractalMind systems coordinate as a larger organism.
| Milestone | Description |
|---|---|
| DAO → execution link | Governance decisions can safely drive remote changes |
| Sub-org autonomy | Independent sub-systems operate under parent oversight |
| Inter-org federation | Multiple organizations collaborate through shared trust and protocol surfaces |
| Operator-grade clients | Better public and internal interfaces for humans steering the system |
Known Gaps
| Gap | Impact | Priority |
|---|---|---|
| Reviewable delivery still uneven across projects | Changes exist but are not always packaged as clean evidence-backed submissions | P0 |
| Shared memory depth is still shallow | State is persistent, but cross-project learning remains immature | P1 |
| Fine-grained capability / security controls | Needed before higher autonomy levels | P1 |
| Distribution bootstrap remains manual | Slows scaling beyond one primary workspace | P1 |
| Public narrative lag | Website and org surfaces can drift behind the real operating model | P1 |
Contributing
The best contributions now are the ones that improve operability:
- make the control loop more reliable
- improve observability
- reduce manual coordination
- produce stronger QA evidence
- align public docs with the system that actually runs