Manual fixtures for loading maggraph.toml, indexing markdown nodes, and trying CLI features.
| Directory | Purpose |
|---|---|
basic/ |
Local-mode config with a two-node sample graph |
lakehouse/ |
Lakehouse-mode config and external-asset node (source URI) |
sync/ |
Leader/follower Git sync configs (bare remote workflow) |
python_agent.py |
Python bindings demo (sync + asyncio) |
Note: Lakehouse HTTP/S3 fetch is stubbed in v0.1 — see
planning/IMPLEMENTATION_STATUS.md.
From the repository root:
cargo test -p maggraph
cargo run -p maggraph-cli -- query --from welcome --depth 2 --config examples/basic/maggraph.toml
cargo run -p maggraph-cli -- ui --config examples/basic/maggraph.toml
cargo run -p maggraph-cli -- scaffold --mcp --skill --config examples/basic/maggraph.toml
cargo run -p maggraph-cli -- --config examples/basic/maggraph.toml --initThe --init flag creates knowledge_graph/ (if missing) and the optional .maggraph/ metadata directory.
examples/basic/
├── maggraph.toml
└── knowledge_graph/
├── welcome.md
└── getting_started.md
examples/lakehouse/
├── maggraph.toml
└── knowledge_graph/
└── customer_churn_q2.md
examples/sync/
├── README.md
├── leader/
└── follower/
root_path in each config is resolved relative to the config file's directory, not the current working directory.
Use GraphIndex::open(resolved.root_path) in Rust to scan all *.md nodes under the graph root.
See sync/README.md for leader push / follower pull steps. Backlog: automated e2e for follower sync init — planning/BACKLOG.md (T-M2).
After maturin develop in python/:
python examples/python_agent.py| Doc | Purpose |
|---|---|
planning/CLI.md |
All CLI commands |
planning/SYNC.md |
Git sync roles |
planning/LAKEHOUSE.md |
Lakehouse mode |
planning/TESTING.md |
Running tests against fixtures |