About io.github.learningloons-hash/lithtrix
io.github.learningloons-hash/lithtrix is an MCP server in the Documentation category: search, browse, memory and blob storage for AI agents. One key at lithtrix.ai, Spark trial free. It has been installed 0 times through Conduid.
Install
claude mcp add io-github-learningloons-hash-lithtrix -- npx -y lithtrix-mcpnpx -y lithtrix-mcpThis server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.
Ask AI
Ask AI about io.github.learningloons-hash/lithtrix
Powered by Claude · Grounded in docs
Security checks
- ·README presentNot checked yet.
- ·License declaredNot checked yet.
- ·Tests presentNot checked yet.
- ·Dependencies pinnedNot checked yet.
- ·No dynamic code executionNot checked yet.
- ·Scoped permissionsNot checked yet.
README
Lithtrix — Agent Infrastructure. Built to Last.
Lithtrix gives AI agents verified APIs for search, Browse (server-side public web for agents with a Sprint / Mission / Deploy pack), memory, document storage, and structured feedback. Agents register themselves in a single API call — no dashboard, no OAuth, no human approval.
Base URL: https://api.lithtrix.ai
Docs: https://docs.lithtrix.ai
Discovery: https://api.lithtrix.ai/v1/capabilities (version 2.5.0 — tier_descriptions, pricing, commons, GET /v1/community, extended _lithtrix.usage + community on authenticated responses)
Pricing: docs/pricing.mdx — Spark trial, Sprint / Mission / Deploy packs, $0.005 per-call rates, auto top-up (Mintlify nav Getting Started → Pricing).
What's Live (Phase 3 + Arc 11)
| Pillar | What it does |
|---|---|
| Search | Credibility-scored web search via GET /v1/search |
| Browse | Server-side public web — buy Sprint (or Mission / Deploy) to unlock (POST /v1/browse, GET /v1/browse/{id}) |
| Memory | Per-agent persistent JSON KV + semantic search via /v1/memory |
| Documents | Binary blob storage, parsing (PDF/DOCX/CSV/XLSX), semantic chunk search via /v1/blobs |
| Feedback | Structured signal on any result via POST /v1/feedback (Arc 11) |
Quickstart (3 calls)
1. Discover
curl https://api.lithtrix.ai/v1/capabilities
2. Register
curl -X POST https://api.lithtrix.ai/v1/register \
-H "Content-Type: application/json" \
-d '{"agent_name":"my-agent","owner_identifier":"you@example.com","agree_to_terms":true}'
Returns { "api_key": "ltx_...", "agent_id": "..." }. The key is shown once — store it immediately.
3. Search
curl "https://api.lithtrix.ai/v1/search?q=Singapore+climate+policy" \
-H "Authorization: Bearer ltx_your_key"
MCP Integration
npx -y lithtrix-mcp
Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"lithtrix": {
"command": "npx",
"args": ["-y", "lithtrix-mcp"],
"env": {
"LITHTRIX_API_KEY": "ltx_your_key_here"
}
}
}
}
Tools exposed include: lithtrix_register, lithtrix_search, lithtrix_browse, lithtrix_feedback, memory (lithtrix_memory_*), and blob tools (lithtrix_blob_*). See MCP and Browse.
Pricing
Spark to start. Sprint, Mission, or Deploy when your agent has work to do.
Credit packs — Spark trial on register ($5, no card), then one-off Sprint ($25), Mission ($50), Deploy ($100) via POST /v1/billing/packs/checkout. Pack grants expire in 90 days (UTC). Metered Search $0.005 and Browse $0.005 per successful call. Paid packs are a private workspace for your agent to search, browse, and remember. Auto top-up: set a threshold — we refill automatically. Full detail: pricing.mdx.
| Pack | Price | Notes |
|---|---|---|
| Spark (trial) | $5 credits on register | Browse not included — buy Sprint (or Mission / Deploy) to unlock Browse |
| Sprint | $25 | One-off credits · 90-day expiry |
| Mission | $50 | One-off credits · 90-day expiry |
| Deploy | $100 | One-off credits · 90-day expiry |
| Volume | Custom | Need more? Get in touch: hello@lithtrix.ai |
Optional referral_agent on register credits the referrer +$0.50 per validated signup (see GET /v1/capabilities referral_rewards). Some older accounts may still show Starter / Pro monthly billing in GET /v1/billing — contact support if you need to migrate.
Credibility Scoring
Every search result includes credibility_score (0.0–1.0):
| Score | Source type |
|---|---|
| 1.0 | .gov domains |
| 0.9 | .edu domains |
| 0.8 | Major news (BBC, Reuters, AP News, NPR…) |
| 0.7 | .org domains |
| 0.5 | All other sources |
Key Endpoints
| Method | Path | Auth | Description |
|---|---|---|---|
| GET | /v1/capabilities |
None | Discovery — version, endpoints, rate limits |
| GET | /.well-known/ai-agent.json |
None | Agent discovery file |
| GET | /v1/guide |
None | Machine-readable agent quickstart |
| POST | /v1/register |
None | Register agent, get API key (agree_to_terms required) |
| POST | /v1/billing/packs/checkout |
Bearer | Buy Sprint / Mission / Deploy credit pack |
| GET | /v1/search |
Bearer | Web search with credibility scoring |
| POST | /v1/browse |
Bearer | Server-side public web (static or dynamic) |
| GET | /v1/browse/{browse_id} |
Bearer | Retrieve logged browse result |
| PUT/GET/DELETE | /v1/memory/{key} |
Bearer | Per-agent JSON memory |
| GET | /v1/memory/search |
Bearer | Semantic memory search |
| PUT/GET/DELETE | /v1/blobs/{id} |
Bearer | Binary blob storage |
| POST | /v1/blobs/{id}/parse |
Bearer | Parse PDF/DOCX/CSV/XLSX |
| GET | /v1/blobs/search |
Bearer | Semantic chunk search |
| POST | /v1/feedback |
Bearer | Structured signal on any result |
| GET | /v1/feedback/stats |
Bearer | Rolling 7d/30d feedback aggregates |
| GET | /v1/me |
Bearer | Agent profile + referral code |
| GET | /v1/usage |
Bearer | Monthly usage + 7-day breakdown |
| POST | /v1/keys/rotate |
Bearer | Rotate API key |
| GET | /health |
None | Health check |
License
MIT
README mirrored from the source repository 17 days ago. The original is authoritative.