Skip to main content
Glama

@sidearmdrm/mcp

MCP server for the Sidearm API — protect media from AI training, detect AI-generated content, and search for stolen work.

This package exposes the Sidearm REST API as Model Context Protocol (MCP) tools that AI agents (Cursor, Claude Desktop, Windsurf, etc.) can call directly.

Quick Start

1. Get an API key

Sign up at sdrm.io and create an API key at sdrm.io/api-keys.

2. Configure your agent

Add to your MCP configuration file:

Cursor (.cursor/mcp.json):

{
  "mcpServers": {
    "sdrm": {
      "command": "npx",
      "args": ["-y", "@sidearmdrm/mcp"],
      "env": {
        "SDRM_API_KEY": "sk_live_..."
      }
    }
  }
}

Claude Desktop (claude_desktop_config.json):

{
  "mcpServers": {
    "sdrm": {
      "command": "npx",
      "args": ["-y", "@sidearmdrm/mcp"],
      "env": {
        "SDRM_API_KEY": "sk_live_..."
      }
    }
  }
}

Windsurf (~/.codeium/windsurf/mcp_config.json):

{
  "mcpServers": {
    "sdrm": {
      "command": "npx",
      "args": ["-y", "@sidearmdrm/mcp"],
      "env": {
        "SDRM_API_KEY": "sk_live_..."
      }
    }
  }
}

That's it. Your agent can now call Sidearm tools.

Related MCP server: GroupDocs.Watermark MCP Server

Tools (27)

Discovery

Tool

Description

list_algorithms

Browse available protection algorithms. Filter by category or media type.

Protection & Extraction

Tool

Description

run_algorithm

Run specific algorithms on media by ID. Returns a job for async processing.

protect_media

Protect media with a preset level (standard/maximum). Auto-selects algorithms.

extract_embeddings

Extract raw embedding vectors from media for downstream similarity search or ML pipelines.

Jobs

Tool

Description

check_job

Poll async job status. Returns progress and results when complete.

Detection

Tool

Description

detect_ai

Detect whether media was AI-generated. Supports image, video, audio, text.

detect_fingerprint

Check if media has been previously registered using fingerprint matching.

detect_membership

Test whether your content was used to train a suspect AI model.

Tool

Description

search_media

Search for similar media across your indexed library.

list_searches

List previous searches on your account.

Media Management

Tool

Description

register_media

Register and index media with optional protection mode (register, search_ready, standard, maximum).

list_media

List media assets in your library (paginated, filterable).

get_media

Get details of a specific media asset including protection status and algorithms applied.

update_media

Update media metadata (e.g., original URL).

delete_media

Permanently delete a media asset and all associated data.

Account, Rights & Billing

Tool

Description

get_account

Get your account details — ID, name, email, credit balance, and plan info.

get_rights

Get C2PA, IPTC, and rights information for a media asset.

get_billing

View credit balance, usage breakdown, per-algorithm costs, and billing portal link.

Provenance & Identification

Tool

Description

get_provenance

Get the full provenance chain for an asset — algorithms, C2PA manifest, membership results.

identify_media

Identify media by its embedded Sidearm fingerprint and extract its C2PA provenance chain.

Shares

Tool

Description

create_share

Create a shareable link for a detection, search, or provenance result (starts private).

get_share

Get a shared result by its share ID.

publish_share

Make a shared result publicly accessible.

Deletion Records

Tool

Description

list_deletions

List deletion records documenting permanently deleted assets.

get_deletion

Get details of a specific deletion record including purged algorithms and storage status.

Documentation

Tool

Description

search_docs

Search the Sidearm API documentation. Returns relevant sections from the full developer reference.

navigate_ui

Look up how to perform actions in the Sidearm dashboard. Returns step-by-step instructions.

Example Agent Conversations

"Protect this image from AI training"

Agent calls protect_media with the image URL, then check_job to get the protected file.

"What algorithms are available for audio?"

Agent calls list_algorithms with media_type: "audio".

"Run Nightshade and Glaze on my artwork"

Agent calls run_algorithm with algorithms: ["nightshade", "glaze"].

"Is this photo AI-generated?"

Agent calls detect_ai with the image, then check_job for results.

"Was my artwork used to train Stable Diffusion?"

Agent calls detect_membership with your media IDs and the model name.

"How do I upload files in the dashboard?"

Agent calls navigate_ui with query: "upload files" for step-by-step instructions.

Environment Variables

Variable

Required

Description

SDRM_API_KEY

Yes

Your Sidearm API key (sk_live_... or sk_test_...)

SDRM_BASE_URL

No

Override the API base URL (default: https://api.sdrm.io)

Development

git clone https://github.com/sidearmDRM/mcp-server.git
cd mcp-server
npm install
npm run build

Test locally with the MCP Inspector:

SDRM_API_KEY=sk_test_... npx @modelcontextprotocol/inspector node dist/index.js

License

MIT

Available Tools

16 tools
check_jobA

Check the status of an asynchronous job (from run_algorithm, protect_media, or detect_ai). Returns status (queued, processing, completed, failed), progress percentage, and result data including download URLs when complete.

Inputs: job_id (required).

delete_mediaA

Permanently delete a registered media asset. Removes storage files, vector embeddings, and all associated metadata. This action cannot be undone.

Inputs: media_id (required).

detect_aiA

Detect whether media content was generated by AI. Supports images, video, audio, and text/PDF. Runs multiple specialized detection models in parallel for the given media type. Returns a job_id — use check_job to poll for results.

Inputs: media_url, media, text, mime, +1 more.

detect_fingerprintA

Detect whether media has been previously registered or seen, using fingerprint matching. Compares against your indexed library at varying depth. Tiers: exact (hash match), quick (perceptual hash), perceptual (visual similarity), compositional (scene structure), full (all tiers). Returns results immediately.

Inputs: media_url, media, tags, tier.

detect_membershipA

Run membership inference to determine whether your protected content was used to train a suspect AI model. Provide content IDs (from your registered media) and the model to test. Methods: pattern (watermark detection), statistical (distribution analysis), combined (both). Returns a job_id — use check_job to poll for results.

Inputs: content_ids (required), suspect_model (required), method, tags.

get_billingB

Get billing and usage events for your account. Returns credit consumption, API call history, and a link to the Stripe customer portal. Filter by date range, event type, or tags.

Inputs: account_id (required), start_date, end_date, type, +1 more.

get_mediaA

Get details of a specific registered media asset by ID. Returns metadata, protection status, applied algorithms, tags, and storage information.

Inputs: media_id (required).

get_rightsB

Get rights and licensing information for a registered media asset. Returns C2PA content credentials, Schema.org structured data, IPTC rights metadata, and TDM-AI protocol declarations. Useful for verifying provenance and license terms.

Inputs: media_id (required).

list_algorithmsA

List available algorithms for media protection, watermarking, and AI content disruption. Returns algorithm IDs, names, supported media types, and descriptions. Use this to discover valid algorithm IDs before calling run_algorithm. Filter by category (open = research algorithms, proprietary = Sidearm bundles) or media_type (image, video, audio, text, pdf, gif).

Inputs: category, media_type.

list_mediaA

List media assets registered to your account. Returns a paginated list with media IDs, types, status, tags, and protection details. Use cursor-based pagination for large libraries.

Inputs: cursor, limit.

list_searchesA

List previous similarity searches performed on your account. Returns a paginated list of past search queries with timestamps and result counts.

Inputs: cursor, limit.

protect_mediaA

Protect media using a curated preset level. Automatically selects the best combination of algorithms for the given media type. Simpler than run_algorithm — just specify standard or maximum protection. Provide either a public media_url, base64 media, or text content. Returns a job_id — use check_job to poll for results.

Inputs: media_url, media, text, mime, +4 more.

register_mediaA

Register and protect media on the Sidearm platform. Modes: register (provenance signing only), search_ready (register + vector indexing), standard (search_ready + watermarks + AI-training poison), maximum (standard + style cloaking + adversarial hardening). Returns the created media object.

Inputs: media_url, media, mode, expires_at, +1 more.

run_algorithmA

Run one or more named algorithms on media. Provide algorithm IDs (from list_algorithms) and either a public media_url or base64-encoded media content. For text, use the text param. Returns a job_id for async processing — use check_job to poll for results. Requires credits.

Inputs: algorithms (required), media_url, media, text, +5 more.

search_mediaB

Search for similar or matching media across the indexed library. Provide a media_url or base64 media to find matches. Tiers: exact (hash match), quick (perceptual hash), perceptual (visual similarity), compositional (scene structure), full (all tiers). Returns results immediately.

Inputs: media_url, media, type, tags, +1 more.

update_mediaB

Update a registered media asset. Currently supports updating the original media URL (e.g., after re-hosting the original file).

Inputs: media_id (required), original_media_url (required).

Tool Schema Changelog

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

Loading tool schema changelog…

    TDQS

    A3.9/5.0

    Scored across 16 tools

    Disambiguation4/5

    Most tools have clearly distinct purposes, such as detect_ai for AI detection, protect_media for protection, and search_media for similarity searches. However, there is some overlap between detect_fingerprint and search_media, as both perform fingerprint matching with similar tier options, which could cause confusion in tool selection.

    Naming Consistency5/5

    All tool names follow a consistent snake_case pattern with clear verb_noun structures, such as check_job, delete_media, and list_algorithms. The naming is predictable and readable throughout the set, with no deviations in style or convention.

    Tool Count4/5

    With 16 tools, the count is slightly high but reasonable for a media protection and detection platform, covering a broad domain including registration, protection, detection, search, and management. It might feel a bit heavy, but each tool appears to serve a specific function without obvious redundancy.

    Completeness5/5

    The tool set provides comprehensive coverage for media management, including CRUD operations (register, get, update, delete), detection (AI, fingerprint, membership), protection (run_algorithm, protect_media), search, and billing. There are no apparent gaps, and the tools support full lifecycle management from registration to deletion.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      C
      maintenance
      Enables capturing evidence, retrieving platform takedown processes, and drafting DMCA/legal notices for content removal, using the Official AI API.
      4
      Apache 2.0
    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables an AI agent to act as a security operations operator by providing tools to monitor camera feeds, analyze scenes, review situations and alarms, acknowledge or resolve alarms, make voice announcements, and consult audit trails and system status.
      -