Cross-domain agricultural intelligence platform
A Note on Documentation Accuracy BijMantra is under active development across genomics, agronomy, AI, interoperability, and full-stack engineering. Exact totals move quickly. For implementation counts and the latest repository snapshot, prefer metrics.json. If you find a discrepancy, please open an issue.
BijMantra is a server-backed agricultural software platform being built to connect workflows that are usually split across separate systems: plant breeding, germplasm and seed banks, seed operations, environmental context, research knowledge, and institutional administration.
It is not intended to be a generic dashboard layer. The goal is to make agricultural decisions less fragmented by keeping genetic, operational, environmental, and institutional context closer together.
At a larger level, BijMantra is also being shaped as an advanced agricultural extension system: a platform that can help shorten the long distance between scientific discovery, institutional validation, and practical field adoption.
The name is literal:
- Bij = seed
- Mantra = utterance or formulation
Snapshot date: 2026-03-23 from metrics.json.
- 250+ application pages across multiple workspaces
- 1500+ API endpoints, including all 201 official BrAPI v2.1 published operations
- Major domain modules and planned module boundaries are in place, but internal cleanup and validation are ongoing
- React 19 + TypeScript + Vite frontend
- FastAPI + PostgreSQL + PostGIS + pgvector + Redis backend stack
- Rust/WASM and Fortran compute components in the wider architecture
- Limited low-connectivity and buffered workflows exist, but the product is not broadly offline-capable today
The exact counts above will drift. This README keeps rounded values on purpose.
Agricultural software becomes weak when it treats each problem as isolated.
BijMantra is hard to build because it sits at the intersection of several difficult constraints:
- Agricultural workflows cross breeding, seed systems, field operations, environment, and knowledge management.
- In agriculture, there is often a 5-15 year gap between research results and broad farmer adoption, driven by validation cycles, limited extension capacity, and risk-aware decision making in the field.
- Data quality is uneven, historical records are inconsistent, and institutional practices vary widely.
- Scientific outputs need provenance, assumptions, and uncertainty, not just values on a screen.
- Some workflows need better resilience under weak connectivity, but the platform still depends on authenticated server-backed flows.
- Hybrid compute makes sense here, but it raises the bar for contracts, validation, and long-term maintenance.
That means the real challenge is not only building software. It is building a system that can carry research insight across institutional, operational, and field-level boundaries in a form people can trust enough to use.
BijMantra remains in preview.
The repository already spans breeding, genomics, gene bank, seed operations, environment, knowledge, and administration, but those areas do not yet share the same maturity or verification depth.
That breadth shows platform direction, not complete production readiness. Some surfaces are already functional, some are being hardened, and some UI mock-up or placeholder views remain as deliberate markers for modules, workflows, or screens still being built.
Important areas still in progress include:
- Deeper validation across modules with uneven operational maturity
- Stronger provenance, contract clarity, and scientific guardrails around intelligence layers
- Better low-connectivity support for selected field workflows without overstating offline capability
- Continued cleanup of broad product surface area into clearer, more stable module boundaries
- Documentation that stays aligned with moving implementation totals
BijMantra is not a small CRUD application. It combines plant breeding science, large biological data models, AI-assisted workflows, and hybrid compute infrastructure.
Because of that scope, the system has already gone through multiple rounds of architectural refactoring. Those changes were driven by real engineering pressure rather than branding:
- scientific features accumulating faster than early boundaries could hold
- module boundaries needing to become more explicit
- AI integration requiring stronger contracts than a simple chat layer
- hybrid Python, Rust/WASM, and Fortran execution needing cleaner separation
Refactoring is treated here as part of building a durable research platform, not as evidence that the effort is directionless.
Docker users: BijMantra uses Podman (rootless, daemonless). Commands are interchangeable.
If you are not a developer, the easiest way to start BijMantra is to use an AI-enabled IDE and let the built-in agent handle setup.
The first startup can take a while because BijMantra may need to install several dependencies and development tools before it can run.
- Install one IDE: Antigravity, VS Code, or Kiro.
- Sign in to your IDE account.
- Open the IDE chat panel and give the agent permission to use the terminal, file system, and Git when prompted.
- Paste this instruction into the chat:
Clone my public repository https://github.com/denishdholaria/bijmantra.git and start BijMantra locally. Install everything required for first-time setup, follow the README, use Podman rather than Docker, and run both the backend and frontend. Expect the first startup to take a while because dependencies and services may need to be installed and initialized.
- After setup finishes, open these local URLs:
- Frontend:
http://localhost:5173 - API docs:
http://localhost:8000/docs
- Frontend:
If the agent asks to install missing tools such as Podman, Python, Node.js, or Make, approve those steps. Dependency installation, container image pulls, and service initialization can take several minutes on first setup.
git clone https://github.com/denishdholaria/bijmantra.git bijmantra && cd bijmantra
bash ./start-bijmantra-app.sh
# Or use the split dev flow
make dev # Start PostgreSQL, Redis, MinIO
make dev-backend # → http://localhost:8000
make dev-frontend # → http://localhost:5173| Service | URL |
|---|---|
| 🌐 Frontend | localhost:5173 |
| 📡 API Docs | localhost:8000/docs |
| 🔌 BrAPI v2.1 | localhost:8000/brapi/v2 |
View Tech Stack Details
| Layer | Technology |
|---|---|
| Frontend | React 19 · TypeScript 5 · Vite · PWA shell |
| Backend | Python 3.13+ · FastAPI · SQLAlchemy · NumPy · SciPy |
| Compute | Rust → WebAssembly · Fortran |
| Data | PostgreSQL · PostGIS · pgvector · Redis · MinIO |
| AI | Multi-provider LLM support · RAG · REEVU |
| Interop | BrAPI v2.1 compatibility work plus repo-specific API surface |
BijMantra can boot from a fresh public clone without live claw runtime folders in the repo root.
- The product repo only keeps public-safe contract material and sanitized bootstrap examples such as
.openclaw.example/and.nemoclaw.example/. - Live claw runtime state, auth stores, device identity, session history, watchdog output, and cron state must stay outside the repo in an operator-managed runtime home such as
~/.bijmantra/runtime/claw. - Operator-only execution surfaces such as the claw compose stack and launcher/watchdog/bridge control scripts are treated as private execution fabric rather than public product contract. In the private repo view, that staged boundary now lives under
ops-private/claw-runtime/.
BijMantra welcomes builders, scientists, and institutions committed to serious agricultural R&D infrastructure. You do not need to be an agriculture expert to contribute.
How to Contribute (Developers & Scientists)
| Action | Impact |
|---|---|
| ⭐Star the repo | Signals community interest to funders |
| 🐛Report a bug | Open an issue — even "this confused me" helps |
| 📝Fix a typo | PRs welcome on docs, comments, UI text |
| Area | Good First Issues |
|---|---|
| Frontend | Component polish, accessibility improvements, page wiring |
| Backend | API endpoint coverage, test coverage expansion |
| Rust/WASM | Genomic algorithm optimization, new compute modules |
| DevOps | CI/CD improvements, container optimization |
# Fork, clone, and start developing
git clone https://github.com/YOUR_USERNAME/bijmantra.git && cd bijmantraorg
make dev && make dev-backend && make dev-frontend📄 CONTRIBUTING.md · CONTRIBUTOR_ENTRY.md
BijMantra's algorithms need validation from people who understand the science. If you work in plant breeding, genomics, agronomy, soil science, or agricultural economics — your review of any module is worth more than a thousand lines of code.
Climate volatility is rewriting agricultural reality. Food systems are already under pressure to produce more with less land, less water, and higher environmental uncertainty.
BijMantra is being built as open agricultural research infrastructure: interoperable where possible, modular, and intended for serious institutional use.
If you want to support the work, collaborate institutionally, or understand sponsorship mechanics, use the dedicated documents rather than treating this README as a funding prospectus.
🚫 Prohibited: Terminator seeds, GURTs, and seed dependency technologies.
📄 LICENSE · COMMERCIAL_LICENSE.md · ETHICAL_USE_POLICY.md
"Agricultural truth emerges at the intersection of disciplines. BijMantra makes that intersection computable."
Bij (बीज) = Seed · Mantra (मन्त्र) = Sacred Utterance — the primordial sound from which the universes were seeded.
The project exists to reduce fragmentation in agricultural data and decisions, and to help serious agricultural knowledge move more effectively from research and institutions into usable practice. It does not assume software alone solves agriculture.
- metrics.json — moving implementation snapshot
- CONTRIBUTING.md — contributor workflow
- CONTRIBUTOR_ENTRY.md — contributor orientation
- ETHICAL_USE_POLICY.md — prohibited and restricted uses
- COMMERCIAL_LICENSE.md — commercial licensing terms
- FUNDING.md — support and backing details
- SPONSORS.md — sponsorship program details
Built for researchers and institutions working toward stronger agricultural decisions.
📧 [email protected] · 🔗 LinkedIn · 🌐 bijmantra.org
