NWT v0.1.0 — MVP

NWT v0.1.0 — MVP

First public release of NeuroWeave Timeline. Implements all six
phases of the MVP spec.

What's in the box

• Core timeline engine — create_event, get_event, list_events, search
• Project evolution graph — typed edges (follows, caused_by, fixes, replaces, extends) + lineage
• CLI — nwt init / log / history / show / search / search-file / search-tag / graph / link / story / explain / rebuild-indices
• MCP server — create_event, search_history, get_project_story, explain_file
• Evolution summaries — deterministic milestone selection (no LLM) + structured project story
• Text-tree graph visualization — Unicode tree with sibling/typed-edge forks

Why standalone

NWT is intentionally independent of NWC. Timeline memory is a general
primitive — useful to engineers, researchers, writers, game designers,
and humans keeping journals. The simpler thing ships first.

See docs/standalone.md.

Quick start

pip install -e .
cd your-project
nwt init
nwt log "Add activation engine" --files activation.py --reason "retrieval was slow"
nwt history
nwt graph
nwt story
nwt explain activation.py

Or with an MCP-aware agent:

{
  "mcpServers": {
    "nwt": { "command": "nwt-mcp" }
  }
}

Storage

Plain JSON under .nwt/. No database, no embeddings, no RAG, no
vector store. Inspectable with cat. Grep-friendly. Git-diff-friendly.

What's NOT in v0.1

By design: vector DB, embeddings, RAG, knowledge-graph reasoning, LLM
planning, autonomous agents, NWC integration. The roadmap is in
docs/roadmap.md.

License

MIT.