Logflare MCP Code Mode
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| LOG_LEVEL | No | trace/debug/info/warn/error/fatal/silent | info |
| LOG_FORMAT | No | json (structured) or pretty (dev) | json |
| VERCEL_TOKEN | Yes | Vercel API token for sandbox creation. | |
| VERCEL_TEAM_ID | Yes | Vercel team ID for sandbox creation. | |
| LOGFLARE_API_KEY | Yes | The Logflare API key used for every execute_read/execute_write call. | |
| VERCEL_PROJECT_ID | Yes | Vercel project ID for sandbox creation. | |
| LOGFLARE_OPENAPI_URL | No | Override the OpenAPI spec URL the search tool loads | https://api.logflare.app/api/openapi |
| LOGFLARE_EXEC_TIMEOUT_MS | No | Per-call budget for the agent's code | 120000 |
| LOGFLARE_SANDBOX_RUNTIME | No | Vercel Sandbox runtime the sandboxes boot from | node24 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| searchA | Search the Logflare Management API's OpenAPI spec by writing JavaScript. Runs in a network-isolated sandbox. Your code is an async arrow function that returns a value; it receives interface OperationInfo { summary?: string; description?: string; deprecated?: boolean; tags?: string[]; parameters?: Array<{ name: string; in: string; required?: boolean; schema?: unknown; description?: string }>; requestBody?: { required?: boolean; content?: Record<string, { schema?: unknown }> }; responses?: Record<string, { description?: string; content?: Record<string, { schema?: unknown }> }>; } declare const spec: { paths: Record<string, Record<string, OperationInfo>> }; Use this to find endpoints and their parameters/request bodies before calling execute_read/execute_write. Output over 100k chars is truncated — return small slices. Example — find endpoints for sources (grouped by path): async () => Object.entries(spec.paths) .filter(([path]) => path.includes('/sources')) .map(([path, ops]) => ({ path, methods: Object.keys(ops) })) Example — inspect one operation's request body and parameters: async () => spec.paths['/api/sources']?.post |
| execute_readA | Read data from the Logflare Management API by writing JavaScript. Only GET requests are allowed — use execute_write for anything that changes state. Example: async () => (await api.request({ method: 'GET', path: '/api/sources' })).data Your code is an async arrow function that returns a value. In scope: declare const api: { request<T = unknown>(options: { method?: "GET" | "POST" | "PUT" | "PATCH" | "DELETE"; // default GET path: string; // e.g. "/api/sources" — see the search tool query?: Record<string, string | number | boolean | undefined>; body?: unknown; // JSON by default contentType?: string; // overrides the default application/json rawBody?: boolean; // send body as-is (no JSON.stringify) }): Promise<{ status: number; ok: boolean; data: T }>; }; Returns a JSON object — result (or error), calledEndpoints, and stdout/stderr — wrapped in an untrusted-data envelope. Use the search tool first to find paths. Access-token, backend, and team endpoints (/api/access-tokens, /api/backends, /api/teams) are blocked — they mint/reveal access tokens, backend connection config, or the account's master API key. |
| execute_writeA | Call the Logflare Management API by writing JavaScript, including state-changing requests (POST/PUT/PATCH/DELETE). Use execute_read for plain reads. Example: async () => (await api.request({ method: 'POST', path: '/api/sources', body: { name: 'my-source' } })).data Your code is an async arrow function that returns a value. In scope: declare const api: { request<T = unknown>(options: { method?: "GET" | "POST" | "PUT" | "PATCH" | "DELETE"; // default GET path: string; // e.g. "/api/sources" — see the search tool query?: Record<string, string | number | boolean | undefined>; body?: unknown; // JSON by default contentType?: string; // overrides the default application/json rawBody?: boolean; // send body as-is (no JSON.stringify) }): Promise<{ status: number; ok: boolean; data: T }>; }; Returns a JSON object — result (or error), calledEndpoints, and stdout/stderr — wrapped in an untrusted-data envelope. Use the search tool first to find paths. Access-token, backend, and team endpoints (/api/access-tokens, /api/backends, /api/teams) are blocked — they mint/reveal access tokens, backend connection config, or the account's master API key. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 3 tools
The three tools have clearly distinct purposes: execute_read for read-only GET requests, execute_write for state-changing requests, and search for exploring the API spec. There is no overlap.
The naming pattern is inconsistent: two tools use the 'execute_' prefix (execute_read, execute_write), while the third is simply 'search', breaking the verb_noun pattern.
Three tools is borderline for a server that aims to cover a full API. While the number is not excessive, it feels minimal, and a typical server might include more dedicated tools instead of relying on generic code execution.
The tool set covers all possible operations on the Logflare Management API: read, write, and spec exploration. Users can perform any CRUD action via execute_read and execute_write, making the surface complete.