Pre-verified answers that save tokens, reduce hallucinations, and cite every source. One API call replaces 5 web searches and 8,000 tokens of parsing.
Product comparisons, software references, business strategy, ERP integrations, compliance rules, and more.
Consumer electronics, computing, software, home, baby, fitness, outdoor, business, finance, energy, compliance, consulting.
No API key required. 1,000 queries/month free. Every unit is CC BY-SA 4.0 licensed.
vs. 3,000-8,000 tokens for web search + parse. Every claim cited with source links.
When an AI agent needs to answer "What are the best wireless earbuds under $150?", it typically:
Total cost: $0.50-$5.00 in compute. Total tokens: 15,000-40,000. Confidence: unknown.
With knowledgelib.io:
GET /api/v1/query?q=best+wireless+earbuds+under+150Total cost: $0.02. Total tokens: ~600. Confidence: 0.88 (scored per published methodology).
The fastest way. Install once, get 6 tools directly in your agent.
Or add to claude_desktop_config.json:
{
"mcpServers": {
"knowledgelib": {
"command": "npx",
"args": ["knowledgelib-mcp"]
}
}
}
Tools: query_knowledge, batch_query, get_unit, list_domains, suggest_question, report_issue
No installation needed. Connect any MCP client to:
Streamable HTTP transport, JSON-RPC 2.0, MCP spec 2025-03-26. Same 6 tools as the stdio server.
# Search
curl https://knowledgelib.io/api/v1/query?q=best+wireless+earbuds
# Batch search (up to 10 queries in one call)
curl -X POST https://knowledgelib.io/api/v1/batch \
-H "Content-Type: application/json" \
-d '{"queries":[{"q":"earbuds"},{"q":"headphones"}]}'
# Get full unit
curl https://knowledgelib.io/api/v1/units/consumer-electronics/audio/wireless-earbuds-under-150/2026.md
# Health check
curl https://knowledgelib.io/api/v1/health
Full OpenAPI spec: /api/v1/openapi.json
Import the Actions schema into any Custom GPT:
6 operations: searchKnowledge, batchSearch, getUnit, suggestTopic, reportIssue, checkHealth.
from langchain_knowledgelib import KnowledgelibRetriever
retriever = KnowledgelibRetriever(
region="US",
entity_type="product_comparison"
)
docs = retriever.invoke("best wireless earbuds")
Community node with 5 operations. Drag-and-drop in any n8n workflow.
| Feature | knowledgelib.io | Web Search | Wikipedia API |
|---|---|---|---|
| Structured metadata | YAML frontmatter + JSON-LD | None | Partial |
| Confidence scores | 0.0-1.0 per unit | None | None |
| Source citations | Every claim cited inline | Page-level only | Footnotes |
| Freshness tracking | Verified date + temporal_validity | Unknown | Edit date only |
| Quality warnings | verified / needs_review / unreliable | None | None |
| Knowledge graph | related_units with typed edges | None | Wikilinks |
| Token efficiency | ~600 tokens per answer | 3,000-8,000 | 1,000-5,000 |
| Agent-specific formats | MCP, REST, LangChain, n8n, GPT Actions | HTML only | REST only |
| ETag caching | 304 Not Modified on repeat access | None | Partial |
| Batch queries | Up to 10 in one call | Sequential only | Limited |
| Type | Count | What It Is |
|---|---|---|
| product_comparison | 418 | Best-of product roundups with decision logic and buy links |
| concept | 336 | Definitions and explanations of terms agents often get wrong |
| software_reference | 239 | Code examples, anti-patterns, debugging guides, decision trees |
| execution_recipe | 202 | Step-by-step implementation plans with gates and deliverables |
| erp_integration | 166 | API capabilities, rate limits, authentication, data mapping |
| agent_prompt | 55 | System prompts for startup pipeline sub-agents |
| assessment | 54 | Structured scoring frameworks for maturity and readiness |
| decision_framework | 35 | Decision trees with trade-offs and anti-patterns |
| benchmark | 28 | Industry benchmarks with segment-specific data |
| rule | 28 | Actionable directives with evidence and scope |
Agents can find knowledgelib.io through any of these standard paths:
knowledgelib.io improves from agent usage:
unreliable in query resultsThe more agents use it, the better it gets. Your queries shape what gets built next.