"Elm" matching MCP tools:

• restore_placeholdersA

  fairquery-llm-data-guardA

  Puts original values back into LLM answers that contain placeholders from a guard_text redact call, so users see real data after redaction.

  MIT

• guard_textA

  fairquery-llm-data-guardA

  Scan text or document URLs for PII and secrets before sharing with an LLM, then redact findings into placeholders, masks, or hashes.

  MIT

• get_run_resultsA

  fairquery-llm-data-guardA

  Poll a guard_text run by run ID to get its status and one page of result rows; continue with nextOffset until finished, without rescanning or re-billing.

  MIT

• list_detectorsA

  fairquery-llm-data-guardA

  Lists supported PII and secret detector types, redaction styles, and billable events to help choose options before redacting text.

  MIT

• execute_codeB

  LLM SandboxB

  Run code in a secure, isolated sandbox across multiple languages, install libraries, and capture generated visualizations for safe execution of LLM-generated code.

  MIT

• get_supported_languagesB

  LLM SandboxB

  Discover which programming languages LLM Sandbox supports for running LLM-generated code in isolated containers.

  MIT

• get_language_detailsC

  LLM SandboxB

  Retrieve details for a programming language supported by LLM Sandbox. Pass the language name to get its information for sandbox code execution.

  MIT

• visionA

  GLM-Vision MCP ServerA

  Convert image files into structured text descriptions and OCR output, enabling text-only LLMs to read and reason about pictures, screenshots, charts, and documents.

  MIT

• get_ai_api_latencyA

  llm-latency-trackerA

  Rank hosted AI inference API providers by measured network latency from a chosen probe region in the last 24 hours. Use it to pick the provider or endpoint closest to your users.

  MIT

• get_provider_latencyA

  llm-latency-trackerA

  Compare one AI provider's latency, uptime, and rank across Germany, USA, Tokyo, and Sao Paulo over the last 24 hours. Use it to see if a provider is slower or unreachable from specific regions.

  MIT

• evm-token-securityA

  The StallB

  Detect honeypots, rug pulls, and scams in any EVM token. Returns a risk score with flags for hidden ownership, mint authority, tax rates, and more across 40+ chains.

  MIT

• get_coding_agent_healthA

  llm-latency-trackerA

  Check if Claude Code or Codex CLI has slowed down, thinks less, or answers worse than its recent baseline. Reports speed, latency, errors, and accuracy drift.

  MIT

• purchase_jsonA

  NeuroDynamic Audio and LLM APIsA

  Purchase narration or LLM text using an owner-funded wallet by providing a unique purchase ID and confirmation; output is saved to your private state directory.

  MIT

• askA

  llm-mcpA

  Send a prompt to a local Qwen model and return the answer text; optionally set a system prompt, model, reasoning mode, or token limit.

  MIT

• get_model_deprecationsA

  llm-latency-trackerA

  Check announced AI model shutdowns and replacements before pinning a version, planning migrations, or verifying if a model ID still works. Filter by provider; optionally include retired models.

  MIT

• evm-nonceA

  The StallB

  Retrieve the confirmed and pending nonce for any EVM wallet address. Use the pending nonce to build new transactions on Ethereum, Base, Polygon, Arbitrum, and Optimism.

  MIT

• budget_statusA

  NeuroDynamic Audio and LLM APIsA

  Check the cumulative client budget, including uncertain purchases reserved across restarts, to track available spending.

  MIT

• get_law_contentA

  e-Gov Law MCP ServerA

  Retrieve Japanese legal statutes from the e-Gov API. Get full law content or summaries for large documents, with JSON or XML formatting options.

  MIT

• authenticated-llm-agentA

  A2A GOSCE AgentsB

  Authenticate requests via JWT and bcrypt, then process JSON inputs through a LangChain-backed Gemini model on Vertex. Rejects unauthenticated calls.

  MIT

• evm-log-eventsA

  The StallB

  Retrieve decoded EVM event logs using contract address and topic signature. Filter by block range and chain to get up to 50 entries per call.

  MIT

• evm-fetch-vaultsA

  @blueprintaix/blueprint-mcpB

  Find and retrieve DeFi vaults across multiple EVM networks, with options to personalize by wallet address and filter by network for targeted yield opportunities.

  -