SuperAPI AI Directory

icogenie-mcp

Generate SVG icons from text descriptions with customizable styles.

Open website

About this tool

IcoGenie MCP (icogenie-mcp) is a Model Context Protocol (MCP) server for IcoGenie, with the goal of enabling AI agents such as Claude to programmatically generate production-ready SVG icons.

  1. Purpose and Installation Add this server to the Claude Desktop configuration to use it (on macOS the path is ~/Library/Application Support/Claude/claude_desktop_config.json). When installed globally or run via npx, the server works automatically.

  2. Authentication On first use, the MCP server will: open a browser for authentication, require approval of access, and save the session token to ~/.icogenie/config.json. Subsequent uses are completed automatically, and the token is shared with the IcoGenie CLI. In CI/CD environments, you can set the ICOGENIE_SESSION_TOKEN environment variable to skip browser authentication.

  3. Available Tools

  • generate_icon: Generates a single icon preview based on a text description. Returns { sessionId, preview, previews, creditsRemaining, sessionData, suggestions }.
  • regenerate_icon: Regenerates a specified icon variant. Returns 4 candidate previews, and confirm_regeneration must be used to confirm the final selection (two-phase process). Returns { candidates, regenToken, creditsRemaining }.
  • confirm_regeneration: Selects one of the 4 candidates to complete regeneration. Cost: free (already included in regenerate_icon). Returns { success, preview, creditsRemaining }.
  • check_credits: Checks the current credit balance. Returns { credits, team, user }.
  • download_icon: Downloads the final SVG + PNG package for an icon or icon set. Cost: single 5 credits, icon sets are 4 credits per icon.
  • normalize_bundle: Generates an AI-enhanced icon list based on a description, used for planning icon sets. Cost: free (rate limited). Returns { bundleId, icons, bundleType, styleRecommendation, reasoning }.
  • generate_bundle: Generates a set of icons based on an icon list. Cost: 1 credit per icon. Returns { bundleId, iconCount, icons: [{ name, description, preview }], pricing, creditsUsed, creditsRemaining }.
  • save_to_library: Saves generated icons to your personal library. Returns { success, libraryItemId }.
  • list_library: Lists icons saved to your personal library. Returns { items: [{ id, name, tags, preview, createdAt }], total, page }.
  • download_from_library: Downloads icons from your personal library in SVG + PNG format. Returns { savedTo } or { base64, filename }.
  • claim_daily_credits: Claims 2 free credits, available once every 24 hours. Returns { credited, creditsRemaining, nextClaimAt }.
  1. Example Workflow Single icon: first check_credits() to check credits; then use generate_icon({ prompt: "notification bell icon", style: "outline" }) to generate, returning sessionId, preview, and remaining credits. If refinement is needed, follow the two-phase process: regenerate_icon({ sessionId: "abc123", index: 0, prompt: "Add a dot indicator" }) to get 4 candidates and a regenToken, then confirm_regeneration({ regenToken: "tok_xyz", selectedIndex: 1 }) confirmation. Finally, use download_icon({ generationId: "abc123", outputPath: "./bell-icon.zip" }) to download the final package. Icon set: first use normalize_bundle({ description: "e-commerce app icons", targetCount: 8 }) for free planning; then use generate_bundle({ icons: [...], style: "outline", bundleId: "bundle_abc" }) to generate (pass in bundleId for easier tracking); finally download_icon({ bundleId: "xyz789", outputPath: "./ecommerce-icons.zip" }) to download.

V. Environment Variables and Related Resources Credits can be purchased at www.icogenie.xyz. Related resources include @icogenie/cli (command-line interface), IcoGenie Web (web application), AI Agent Docs (machine-readable documentation), and Developer Docs (integration guide). The official documentation does not provide further explanation regarding the specific scenarios for choosing MCP versus CLI or Web.

VI. Applicable Scenarios and Limitations It is suitable for batch-generating and iterating SVG icons through conversational descriptions in AI clients that support MCP, and for managing a personal icon library. In terms of limitations: some tools consume credits (downloading a single icon costs 5 credits, an icon set costs 4 credits per icon, and generating an icon set costs 1 credit per icon); regeneration and confirmation use a two-stage process; free credits can be claimed 2 per 24 hours per day; first-time use requires browser authentication, and CI/CD environments need to set a session token environment variable. Other feature details, platform support scope, pricing, and results are not yet provided in the available information.