How to Contribute
1
Fork & Clone
Fork the repo on GitHub and clone it locally. git clone https://github.com/YOUR_USERNAME/graphista.git
2
Pick an Issue
Browse open issues below. Look for good-first-issue labels if you're new. Bounty issues have token rewards.
3
Branch & Build
Create a feature branch: git checkout -b feat/your-feature โ Write clean code, add tests, update docs.
4
Submit PR
Push your branch, open a PR against main. Describe what you changed and why. We review within 24-48h.
Open Issues
12 Open#142 Multi-agent shared memory graphs
feature500K tokens
Implement shared memory graphs that allow multiple AI agents to read/write to the same graph with conflict resolution. Needs CRDT-based merge strategy for concurrent writes.
#139 GraphQL API for external queries
featurecore
Expose the memory graph via a GraphQL endpoint so external apps can query Graphista without using the Python SDK. Schema design needed.
#137 Memory decay and relevance scoring
feature250K tokens
Implement time-based decay on memory nodes. Frequently accessed nodes stay strong, rarely accessed ones fade. Configurable decay curves.
#134 Edge weight visualization broken on large graphs
buggood first issue
When graph exceeds 500 nodes, edge weight rendering causes frame drops. Need to implement LOD (level of detail) culling for distant edges.
#131 Add Rust bindings for graph engine
featureperformance
Port the core graph traversal engine to Rust via PyO3 for 10-50x performance improvement on large graph queries. Python API stays the same.
#128 Improve ingestion docs with examples
docsgood first issue
Current ingestion docs lack real-world examples. Add cookbook-style guides for: chat logs, PDF documents, API responses, and code repositories.
#125 Solana on-chain memory anchoring
feature1M tokens
Anchor memory graph hashes on-chain for verifiable, tamper-proof memory. Merkle tree of node states committed to Solana at configurable intervals.
#122 Streaming ingestion pipeline
featurecore
Support real-time streaming ingestion via WebSocket. Currently ingestion is batch-only. Need async pipeline with backpressure handling.
Pull Requests
0 OpenNo pull requests yet
Be the first to contribute โ pick an issue above and submit a PR
Autonomous Review Agents
CodeReview-Agent
Scans PRs for bugs, style, test coverage
ACTIVE
Reviewed: 0
Issues found: 0
HealthCheck-Agent
Monitors codebase for regressions
ACTIVE
Files scanned: 0
Health: 100%
PathMapper-Agent
Auto-generates implementation guides for issues
ACTIVE
Issues mapped: 0
Paths generated: 0
Agent Activity Log