Privacy Policy
Last updated: March 9, 2026
Overview
Synaptic is a local-first tool. Your data stays on your machine by default. We do not operate servers that collect, store, or process your data. This policy explains what data Synaptic handles and where it goes.
Data Storage
All context entries, embeddings, rules, and search indexes are stored in a local SQLite database on your machine at ~/.synaptic/. This data never leaves your machine unless you explicitly enable cross-machine sync.
Synaptic does not have access to your files, source code, or git repositories beyond what Claude Code's MCP protocol provides during a session.
Cross-Machine Sync
If you opt in to cross-machine sync, Synaptic pushes context entries to a private GitHub repository that you own. Data flows exclusively between your machines and your own GitHub account via the gh CLI. No data passes through HYPERLYNQ servers.
LLM Fact Extraction (Optional)
Synaptic includes an optional feature that sends small, structured snippets of tool output to the Anthropic API (Claude Haiku) for intelligent fact extraction. This feature:
- Only activates if an API token is available (your existing Claude Code session token or an
ANTHROPIC_API_KEYenvironment variable) - Sends only pattern-matched snippets (database schemas, route patterns, error messages) — not your full source code or conversation
- Redacts credential values, environment variable values, and secrets before transmission
- Data is sent to Anthropic's API at
api.anthropic.comand is subject to Anthropic's privacy policy - Falls back gracefully if no token is available — Synaptic works fully without this feature
Embedding Model
Synaptic uses a local Hugging Face model (all-MiniLM-L6-v2) to generate 384-dimensional embeddings for semantic search. The model runs entirely on your machine. No text is sent to Hugging Face or any external service for embedding generation.
Analytics & Telemetry
Synaptic does not include any analytics, telemetry, crash reporting, or usage tracking. We do not know how many users we have, how often Synaptic is used, or what features are popular.
Website
The synaptic.dev website is a static site hosted on Vercel. Vercel may collect standard web server logs (IP addresses, request timestamps) as described in Vercel's privacy policy. We do not add any additional tracking scripts, cookies, or analytics to the website.
npm Package
Synaptic is distributed via npm. When you install the package, npm may collect usage data as described in npm's privacy policy. Synaptic itself does not execute any install scripts that collect or transmit data.
Third-Party Services
Summary of external services Synaptic may interact with:
| Service | When | Data Sent |
|---|---|---|
| Anthropic API | Optional LLM extraction (opt-in via token) | Redacted tool output snippets only |
| GitHub | Optional cross-machine sync (opt-in) | Context entries to your own private repo |
| npm | Package installation | Standard npm install metadata |
Changes to This Policy
We may update this privacy policy from time to time. Changes will be posted on this page with an updated revision date. Significant changes will be noted in the GitHub release notes.
Contact
Questions about this policy? Contact us at hyperlynq@gmail.com.