Search
Semantic search across your extracted knowledge.
The search endpoint lets you query your extracted knowledge using natural language. It uses hybrid keyword + vector search powered by Typesense to deliver accurate, relevant results from all the knowledge you have extracted.
Search uses hybrid keyword + vector search for best results. This combines exact keyword matching with semantic understanding, so queries like "how much does it cost" will match knowledge items about pricing even if they do not contain the exact phrase.
Search Knowledge
/v1/searchx-api-keySearch across all knowledge items extracted under your API key. Returns ranked results with relevance scores.
Request Parameters
querystringrequiredThe natural language search query. Can be a question, phrase, or keywords. The search engine handles both exact keyword matches and semantic similarity.
limitnumberMaximum number of results to return. Minimum: 1, Maximum: 100.
categorystringFilter results to a specific knowledge category. Useful when you know the type of information you are looking for. Accepted values: PRODUCT, FEATURE, PRICING, FAQ, SUPPORT, COMPANY, LEGAL, OTHER.
Response
resultsarrayArray of matching knowledge items, ranked by relevance score. Each item contains:
results[].titlestringThe title of the knowledge item.
results[].descriptionstringA short summary of the knowledge item.
results[].contentstringThe full text content of the knowledge item.
results[].categorystringThe category of the knowledge item (e.g., PRODUCT, FEATURE, PRICING).
results[].sourcestringThe URL the knowledge item was extracted from.
results[].scorenumberRelevance score between 0 and 1. Higher scores indicate better matches.
Example
Example Response
{
"results": [
{
"title": "GitHub Integration",
"description": "Connect your GitHub repositories to automatically link issues, PRs, and commits.",
"content": "The GitHub integration allows you to link issues to pull requests, auto-close issues when PRs are merged, and see commit history directly in your project tracker. Setup takes less than a minute — just authorize the GitHub app and select your repositories.",
"category": "FEATURE",
"source": "https://linear.app/integrations",
"score": 0.96
},
{
"title": "Slack Integration",
"description": "Create and manage issues directly from Slack with two-way sync.",
"content": "The Slack integration lets your team create issues from any message, get notified about updates in channels, and manage issue status without leaving Slack. Supports slash commands, message actions, and automated notifications.",
"category": "FEATURE",
"source": "https://linear.app/integrations",
"score": 0.93
},
{
"title": "Figma Integration",
"description": "Embed Figma designs directly in issues for design-to-development handoff.",
"content": "Paste any Figma link into an issue and it will automatically embed a live preview. Designers and developers can reference the latest designs without switching tools.",
"category": "FEATURE",
"source": "https://linear.app/integrations",
"score": 0.89
}
]
}Filtering by Category
Available Categories
| Category | Description | Example queries |
|---|---|---|
PRODUCT | Core product information | "What does this product do?" |
FEATURE | Specific features and capabilities | "Does it support SSO?" |
PRICING | Plans, tiers, and billing details | "How much does the pro plan cost?" |
FAQ | Frequently asked questions | "How do I reset my password?" |
SUPPORT | Help articles and troubleshooting | "How to contact support?" |
COMPANY | About the company, team, and mission | "Who founded the company?" |
LEGAL | Terms of service, privacy, compliance | "What is the data retention policy?" |
OTHER | Other content | General queries |
For the most comprehensive results, omit the category parameter. The search engine will rank all relevant items regardless of category. Use category filtering only when you need results from a specific type.