# HikerAPI [Health: Active]

**Category:** 🌐 Social Media  
**Repository:** https://github.com/subzeroid/hikerapi-mcp  
**GitHub Stars:** 8  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/hikerapi

## Description
Instagram data via HikerAPI: profiles, posts, reels, stories, comments, hashtags, locations.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "hikerapi": {
    "command": "npx",
    "args": ["-y","hikerapi-mcp"],
    "env": {
      "HIKERAPI_KEY": ""
    }
  }
}
```

**Requires environment variables:** `HIKERAPI_KEY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What HikerAPI MCP server does

The HikerAPI MCP server connects an MCP-compatible AI assistant to HikerAPI, an API for reading Instagram data. It covers profile lookups, user media, followers, posts, comments, likers, account and reel search, hashtags, locations, stories, highlights, shared links, and audio tracks.

The server is read-only. Each exposed MCP tool corresponds to a HikerAPI HTTP GET operation, and the tool name reflects the endpoint path. For example, an endpoint for finding a user by username becomes a tool named `get_v2_user_by_username`. Returned response bodies are passed back as JSON text.

This belongs in the 🌐 Social Media category and is suited to assistants that need structured Instagram lookups rather than browser automation or write access.

## How it works

At startup, the HikerAPI MCP server downloads the live OpenAPI document from the configured API host. It uses that document to build MCP tool definitions, including argument schemas for endpoint path and query parameters. The server then communicates with the assistant over stdio and sends HTTPS requests to HikerAPI.

The default configuration exposes a curated core collection of roughly 45 tools. These select one endpoint for each common task instead of presenting every version variant to the model. Setting `HIKERAPI_TOOLS=all` exposes more than 100 non-deprecated endpoints. Deprecated operations and the `Legacy` and `System` groups remain excluded. API errors are returned as tool errors containing the HTTP status and response body.

## Setup and configuration

Install and run the HikerAPI MCP server with `npx -y hikerapi-mcp`, then provide an API key through `HIKERAPI_KEY`. The key is sent in the `x-access-key` request header. The README documents configuration examples for Claude Desktop, Claude Code, Cursor, Windsurf, Zed, and OpenAI Codex.

`HIKERAPI_URL` changes the API base URL, while `HIKERAPI_SPEC_URL` controls where the OpenAPI document is loaded from. `HIKERAPI_TOOLS` selects the core or all mode. `HIKERAPI_TAGS` can restrict tools to selected comma-separated tags, and `HIKERAPI_EXCLUDE_TAGS` adds tags to the default exclusions. Timeout and response-size limits can be adjusted with the documented `*_TIMEOUT_MS` and `*_MAX_*_BYTES` variables.

Only use a custom API host when it is an intended self-hosted or proxied HikerAPI service. The server warns if the host is outside the two recognized HikerAPI hostnames because the access key will be sent there.

## Tools and capabilities

The HikerAPI MCP server can expose tools for:

- Looking up Instagram profiles and user media
- Reading post details, comments, and liker data
- Searching accounts and reels
- Retrieving popular or recent hashtag media
- Reading location details and recent location media
- Fetching stories, highlights, and shared-link data
- Looking up audio tracks

Use `tools/list` over MCP to inspect the current generated tool set and parameter schemas, since the available operations come from the live specification.

## Limitations and notes

The server only exposes GET operations, so it does not create, edit, delete, like, follow, or otherwise mutate Instagram data. API availability and response content depend on HikerAPI. The default core selection may omit less common endpoints; use all mode or tag filters when the required operation is not included.

The OpenAPI specification is fetched during startup, with configurable retry, timeout, and size limits. HikerAPI request responses are subject to the configured maximum response size, which defaults to 10 MB. The project is licensed under MIT.

_Full upstream README: https://allmcps.com/mcp/hikerapi/readme_