Current release: v0.1.0 — a tamper-evident, verifiable ledger for AI agent cognition.
Target: Q3 2026
The checkpoint layer is established. v0.2 closes the loop by making the system self-correcting — agents that encounter transient failures recover automatically without human intervention.
| Feature | Description |
|---|---|
korg autoheal |
Autonomous recovery loop: detect failure → identify root cause → rewind to last clean checkpoint → retry with corrected strategy |
| Incremental checkpoints | Background checkpoint compression — only delta since last checkpoint is stored |
| Recovery confidence scoring | Each autoheal attempt is scored; low-confidence recoveries escalate to human approval |
| Doom-loop circuit breaker | Hard limit on recursive recovery attempts with structured escalation |
Target: Q4 2026
Korg's governance kernel should be usable from any language, not just Rust CLIs.
| Feature | Description |
|---|---|
Python SDK (pip install korg) |
korg.Session, korg.rewind(), korg.fork() — full runtime access from Python |
| JavaScript/TypeScript SDK | Node.js + Deno compatible; works with LangChain, Vercel AI SDK |
| gRPC API | Language-agnostic wire protocol for embedding Korg in existing agent frameworks |
| REST API | POST /api/campaign, GET /api/ledger/:seq, POST /api/rewind |
Target: Q1 2027
| Feature | Description |
|---|---|
| Custom capability registration | Agents can register domain-specific capabilities (e.g. database_query, image_generate) with full ledger governance |
| Capability marketplace | Community-contributed capability plugins, version-locked and hash-verified |
| Sandboxed capability execution | Each capability runs in an isolated workspace with blast-radius bounds |
- Cloud sync — share sessions and ledger state across team members
- Distributed swarm — multi-machine campaign execution with network-aware HLC
- Formal verification — TLA+ spec for the core ledger invariants
See CONTRIBUTING.md for architecture invariants and how to propose new capabilities.
Open a discussion if you want to influence the roadmap — especially for the SDK design in v0.3.