by Peace Thabiwa — SAGEWORKS AI
MindScript Ledger is the memory fabric for the MindScript ecosystem.
Instead of relying on raw chat history, it stores:
- normalized prompts (inputs)
- logic and stage metadata
- pattern hashes
- temporal order
- threads of interaction
- snapshots and deltas
This lets you:
- reconstruct past sessions deterministically
- recall patterns instead of keywords
- replay state transitions
- share memory across different AI models (GPT, Gemini, etc.)
2-OVERVIEW.md— conceptual overview3-spec/— full ledger model specification4-schemas/— JSON schemas for entries, threads, patterns, snapshots5-storage/— storage layout + compaction rules6-engine/— ingestion, state machine, reconstruction docs7-query-language/— MSQL (MindScript Query Language)8-integrations/— how this binds to MindScript Core + Templates9-examples/— concrete example files10-roadmap.md— future plans
- v0.1 — Specification + schemas
- v0.2 — Reference implementations (planned)
- v1.0 — Production-ready engine (future)
MindScript repos:
mindscript-core— language spec & lawsmindscript-templates— multimodal templatesmindscript-ledger— this repo, temporal memory layer