Provides tools for Akamai edge diagnostics, debugging, and network analysis using Akamai APIs.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@mcp-connector-for-akamai List available edge server locations"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Disclaimer: This is an independent, unofficial, community-driven project. It is not affiliated with, endorsed by, sponsored by, or in any way officially connected to Akamai Technologies, Inc. The name "Akamai" and any related trademarks, service marks, or logos are the property of their respective owners and are used here solely for descriptive purposes to indicate compatibility. Use of these names does not imply any affiliation or endorsement. This software is provided "as is", without warranty of any kind. The authors assume no liability for any damages arising from the use of this software.
An unofficial MCP (Model Context Protocol) server that connects AI assistants to Akamai APIs. Provides tools for edge diagnostics, debugging, and network analysis.
Link to NPM: https://npmjs.com/package/mcp-connector-for-akamai
Tool | Description |
| Decode Akamai error reference codes into detailed error info |
| Fetch a URL from an Akamai edge server with debug headers |
| HTTP error statistics for a URL or CP code |
| Verify if an IP is Akamai edge and get geolocation |
| Geolocate up to 10 IP addresses |
| DNS lookup from an Akamai edge server |
| MTR traceroute from an Akamai edge server |
| Search edge server logs by IP, time range, and filters |
| List available Akamai edge server locations |
| Decode an Akamaized URL (ARL) into components |
| Verify up to 10 IPs as Akamai edge IPs |
| Comprehensive URL health check (cURL + dig + MTR) |
| Diagnose connectivity issues (GREP + cURL + MTR) |
| Diagnose content/cache issues (edge vs origin comparison) |
Related MCP server: mcp-browser-kit
Node.js 18+ or Bun
Akamai API credentials with Edge Diagnostics API access
bun install
bun run buildVariable | Required | Description |
| Yes | Akamai EdgeGrid client secret |
| Yes | Akamai EdgeGrid client token |
| Yes | Akamai EdgeGrid access token |
| Yes | Akamai API hostname (e.g. |
| No | Account switch key for multi-account setups |
Add to your .vscode/mcp.json:
{
"inputs": [
{ "id": "akamai-client-secret", "type": "promptString", "description": "Akamai EdgeGrid client secret", "password": true },
{ "id": "akamai-client-token", "type": "promptString", "description": "Akamai EdgeGrid client token", "password": true },
{ "id": "akamai-access-token", "type": "promptString", "description": "Akamai EdgeGrid access token", "password": true },
{ "id": "akamai-host", "type": "promptString", "description": "Akamai API hostname (e.g. akab-xxx.luna.akamaiapis.net)" }
],
"servers": {
"mcp-connector-for-akamai": {
"type": "stdio",
"command": "npx",
"args": ["-y", "mcp-connector-for-akamai"],
"env": {
"AKAMAI_CLIENT_SECRET": "${input:akamai-client-secret}",
"AKAMAI_CLIENT_TOKEN": "${input:akamai-client-token}",
"AKAMAI_ACCESS_TOKEN": "${input:akamai-access-token}",
"AKAMAI_HOST": "${input:akamai-host}"
}
}
}
}Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"mcp-connector-for-akamai": {
"command": "npx",
"args": ["-y", "mcp-connector-for-akamai"],
"env": {
"AKAMAI_CLIENT_SECRET": "...",
"AKAMAI_CLIENT_TOKEN": "...",
"AKAMAI_ACCESS_TOKEN": "...",
"AKAMAI_HOST": "..."
}
}
}
}Generate the API client: add the OpenAPI spec URL to generate-clients.sh (in the parent workspace) and run it
Create a wrapper in src/akamai/apis/<api-name>/
Add tool registration files under src/tools/<api-name>/
Register the tools in src/tools/index.ts
src/
├── index.ts # Entry point (McpServer + STDIO transport)
├── akamai/
│ ├── client.ts # Namespaced client hub (edgeDiagnostics, future APIs)
│ ├── customFetch.ts # EdgeGrid-authenticated fetch
│ ├── generateAkamaiToken.ts # EdgeGrid auth token generation
│ └── apis/
│ └── edge-diagnostics/
│ ├── index.ts # Client factory & re-exports
│ └── v1/ # Auto-generated API client
└── tools/
├── index.ts # Tool registration hub
├── utils.ts # Polling, formatting helpers
└── edge-diagnostics/ # One file per tool or tool groupRuns Akamai connectivity diagnostics for a URL — simultaneously executes GREP (edge logs), cURL (content fetch), and MTR (traceroute). Useful for diagnosing slow downloads and high response times. This is an async operation.
Inputs: url (required), port, clientIp, ipVersion, +5 more.
Runs Akamai content problem diagnostics for a URL — fetches content via cURL from both edge and origin to compare responses. Useful for diagnosing cache inconsistencies, stale content, and origin errors. This is an async operation.
Inputs: url (required), ipVersion, spoofEdgeIp, edgeLocationId, +2 more.
Fetches a URL from an Akamai edge server, returning full HTTP response headers (including Akamai debug/Pragma headers) and body. Useful for debugging what the edge is serving, cache status, and response timing.
Inputs: url (required), edgeIp, ipVersion, spoofEdgeIp, +3 more.
Runs a DNS dig command from an Akamai edge server, returning DNS records for a hostname. Useful for debugging DNS resolution and propagation issues.
Inputs: hostname (required), queryType (required), edgeIp, isGtmHostname, +1 more.
Returns HTTP error statistics (status code distribution, error percentages) for a URL or CP code. Shows edge errors and/or origin errors over the last 24-48 hours.
Inputs: url, cpCode, delivery, errorType.
Searches edge server logs by edge IP, time range, and filters (CP codes, hostnames, status codes, client IPs, user agents). This is an async operation that polls until logs are available.
Inputs: end (required), start (required), edgeIp (required), logType (required), +5 more.
Lists available Akamai edge server locations. Returns location IDs that can be used as edgeLocationId in other tools (curl, dig, MTR, etc.).
Inputs: no named parameters.
Locates up to 10 IP addresses, returning geolocation data for each (city, country, region, ASN, network).
Inputs: ipAddresses (required).
Runs an MTR (My Traceroute) network diagnostic from an Akamai edge server to a destination IP or hostname. Shows network path, latency per hop, and packet loss. Useful for diagnosing connectivity issues.
Inputs: packetType (required), destination (required), destinationType (required), port, +6 more.
Translates an Akamaized URL (ARL) into its component parts: CP code, serial number, TTL, origin server, and other cache key metadata.
Inputs: url (required).
Translates an Akamai error reference code (e.g. "9.6f64d440.1318965461.2f2b078") into human-readable error details including logs, HTTP status codes, and request/response metadata. This is an async operation that polls until complete.
Inputs: errorCode (required), traceForwardLogs.
Runs a comprehensive health check on a URL from the Akamai edge, combining cURL, dig, and optionally MTR. This is an async operation that polls until complete.
Inputs: url (required), port, ipVersion, queryType, +6 more.
Verifies whether up to 10 IP addresses belong to the Akamai edge network.
Inputs: ipAddresses (required).
Verifies whether an IP address belongs to the Akamai edge network and returns its geolocation details (city, country, region, ASN, network).
Inputs: ipAddress (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 14 tools
Several diagnostic tools overlap: url_health_check, connectivity_problems, and content_problems all perform edge-side diagnostics with cURL/dig/MTR, though they target different symptoms. Additionally, verify_edge_ip and verify_locate_ip serve similar verification purposes. However, descriptions are detailed enough to differentiate most tools.
Names follow a mix of verb_noun (list_edge_locations, translate_error_string) and noun phrases (url_health_check, connectivity_problems). The verb_from_edge pattern for curl, dig, mtr is consistent for those tools, but overall the set lacks a uniform naming convention.
With 14 tools, the set is appropriately sized for an Akamai diagnostics server. Each tool addresses a specific network or content debugging need without excessive redundancy.
The server covers core diagnostic workflows: URL translation, IP verification/location, edge commands, error stats, log search, and problem detection. Missing is a location details getter and maybe a direct 'get_edge_location' but most operations are covered. Minor gaps only.
MCP server for AI agents to plan, verify, and deploy Cloudflare-native apps.
MCP server for OpenAI API (chat completions, image generation, embeddings) via AceDataCloud
An MCP server that integrates with Discord to provide AI-powered features.
MCP server that lets your AI assistant build and deploy apps, hosting each one at a live URL.
