Rta-Smriti Brain
Stop re-explaining your project to every new AI chat.
3 followers
Stop re-explaining your project to every new AI chat.
3 followers
Rta-Smriti gives every software project a private local brain. It indexes repo structure, remembers decisions and handoffs, labels evidence, and builds focused context packs for Codex, Claude Code, Cursor, or any agent. Local, open source, no telemetry.
Rta-Smriti Brain has reached v1.0.4-alpha. The project now goes beyond searchable repository memory: it reconciles repository evidence, durable decisions, checkpoints, session continuity, and provenance into an inspectable local Project Reality view.
The next test is operator activation. If you use an AI coding agent across a long-lived project, try one non-sensitive project and tell me whether you reached a useful context pack or where setup failed.
Release: https://github.com/sulabhdubey/r...
Small launch-day update: v0.4.0-alpha is now live as a verified cross-platform release.
The fastest path is now one command: point Rta-Smriti at a repository, and it creates the brain, indexes the project, starts incremental sync, and opens the local operator console. The release also adds structured continuation checkpoints, provenance, hybrid retrieval, Tree-sitter/LSP adapters, retrieval diagnostics, and an evidence-aware Action Gate.
I would value one hard test more than generic praise: try it on a non-sensitive repository and tell me exactly where onboarding, retrieval, or the dashboard becomes confusing.
Release: https://github.com/sulabhdubey/rta-smriti-brain/releases/tag/v0.4.0-alpha