> ## Documentation Index
> Fetch the complete documentation index at: https://docs.reposeek.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# HTTP and curl

> Call Reposeek from any HTTP client.

# HTTP and curl

Use any HTTP client that can send JSON and Bearer auth.

```bash theme={null}
curl -sS -X POST "https://api.reposeek.ai/v1/search" \
  -H "Authorization: Bearer ${REPOSEEK_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{"query":"local AI assistant that remembers screen activity","limit":5,"filters":{"stars":{"min":1000},"forks":{"min":10},"license":["MIT","Apache-2.0"]}}'
```

<a href="https://reposeek.ai/dashboard/playground?endpoint=search&query=local%20AI%20assistant%20that%20remembers%20screen%20activity&limit=5&min_stars=1000&min_forks=10&license=MIT&license=Apache-2.0">Explore in API Playground</a>

Sign in to run this request from your dashboard.

Successful responses contain top-level `request_id` and `results`; each result includes `stars` and `forks`. Use `X-Request-Id` for support/debug correlation and `response.headers.get("Server-Timing")` in clients that need route-level latency.

```bash theme={null}
curl -sS -X POST "https://api.reposeek.ai/v1/search" \
  -H "Authorization: Bearer ${REPOSEEK_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{"query":"react data table","limit":3}' | jq
```