Skip to main content
Glama

image-token-saver

image-token-saver is a deterministic image reducer. It has no LLM inside the reducer. It runs as a local MCP server on stdio. An agent calls it before the agent sends an image to a vision model. The tool reduces the input image tokens.

Licence

This project uses the MIT licence. See LICENSE.

Related MCP server: openmedia-mcp

Versions

  • Node.js 24.22.3

  • TypeScript 7.0.2

  • @modelcontextprotocol/sdk 1.32.2

  • sharp 0.35.5

The project is ESM. Tests use node:test.

How to run (later issues)

The library and the MCP server come in later issues. When they exist, you will start the server with one command on stdio. This README will show that command.

Run the local checks before each commit:

scripts/precommit.sh

Sample renders

Local sample renders are not committed. The .gitignore file ignores the local sample directories. Tests use synthetic fixtures only.

Documents

  • docs/PRD.md: product requirements.

  • docs/overview.md: project overview.

  • GLOSSARY.md: terms.

  • CLAUDE.md: rules for coding agents.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      A
      quality
      Not graded
      maintenance
      Enables AI agents to analyze images through vision AI providers (Gemini, OpenAI, Claude), performing tasks like image description, object detection with bounding boxes, region-specific analysis, and precise color extraction without consuming context window with raw pixels.
      4
      -
    • A
      license
      A
      quality
      B
      maintenance
      Enables AI agents to perform deterministic, non-generative image transformations—straightening, cropping, masking, layering, color adjustments, and encoding—as reproducible recipes while preserving immutable originals.
      13
      MIT
    • A
      license
      A
      quality
      C
      maintenance
      Enables AI agents to perform local image operations such as circle cropping, cropping, resizing, compressing, converting, and inspecting images without uploading files or making network calls.
      7
      85 npm
      MIT