Installation
Spine ships as the synaptixs-spine package. The CLI is orchestrator; the MCP
server is orchestrator-mcp. Requires Python 3.12+.
Core
pip install synaptixs-spineThis gives you deterministic, no-LLM comprehension across all 14 language
front-ends and the full orchestrator CLI (55 commands, 7 areas).
The full install
pip install 'synaptixs-spine[all]'[all] is the recommended install for the assistant plugin: every language
front-end, the MCP server, and PDF + Word/Excel ingestion.
Targeted extras
Install only what you need:
| Extra | Adds |
|---|---|
all | every front-end + MCP server + doc/office ingestion (recommended) |
languages | all 14 language front-ends |
sdlc | run the generated tests (the sdlc feature/run path) |
mcp | the orchestrator-mcp assistant plugin server |
docs | .pdf ingestion |
office | .docx, .xlsx ingestion |
sql | SQL data-layer comprehension + codegen |
go | Go language support |
rust | Rust language support (comprehension + Cargo-aware codegen; also needs Cargo, rustc and Rustfmt at the repository’s declared MSRV or newer, and Clippy when the repository configures it) |
php | PHP language support (comprehension + codegen via Composer or a pinned PHPUnit PHAR) |
perl | Perl language support (comprehension + codegen via perl -c, Perl::Critic, prove) |
kotlin | Kotlin language support (comprehension + call graph; codegen via a Gradle-aware test runner for Kotlin/JVM and Android) |
clang | optional C/C++ semantic pass — wheel-bundled libclang resolves member calls the CST cannot, adding edges only between symbols already in the graph; included in [all] |
media | opt-in image OCR (Tesseract) + media transcripts — off by default |
asr | opt-in audio/video transcription (local Whisper) |
otel | OpenTelemetry export |
Verify
orchestrator --version # → 3.57.1orchestrator understand .