Skip to main content
Glama

MCP Tool-Poisoning, Prompt-Injection & 2026 EU AI Act Auditor MCP

Server Details

Scan external MCP servers for tool-poisoning, hidden Unicode injection & EU AI Act Annex III.

If you are the author of this connector, you can claim ownership. Claimed connector authors can inspect health checks, view analytics, and manage their listing.
Status
Healthy
Uptime
100.0% over the last 24 hours
Last Tested
Transport
Streamable HTTP · MCP 2025-03-26
URL
Repository
BeakDo/defi-customs-mcp-hub
GitHub Stars
0
ChatGPT
–
Claude
–

TDQS

A3.5/5.0

Scored across 6 tools

Disambiguation4/5

The three audit tools are mostly distinguishable by target: MCP schema, single package CVE, and batch dependency CVE. However, the single-package and batch SBOM CVE audits overlap enough that an agent could reasonably hesitate between them, and the remaining energy/patent/clinical tools are clearly separate.

Naming Consistency4/5

All names are snake_case and generally follow a verb_noun or action_target pattern, such as audit_package_cve_and_cisa_kev, optimize_compute_energy_grid_dispatch, search_patent_academic_prior_art, and track_biotech_clinical_trials. Minor deviations include batch_sbom_cve_audit_x402, which leads with a noun modifier and appends a payment suffix, and several unusually long names.

Tool Count4/5

Six tools is a reasonable count and avoids overwhelming the caller. The set is slightly awkward because the tools span disconnected domains—security auditing, CVE lookup, energy dispatch, patents, and clinical trials—so not every tool clearly serves the server's advertised auditing purpose.

Completeness3/5

The auditing surface covers MCP schema inspection, single-package CVE checks, and batch dependency checks, but lacks obvious supporting operations such as remediation guidance, report export, or batch MCP auditing. The non-auditing tools each appear as isolated one-off capabilities, making full lifecycle coverage hard to assess and unlikely.

Available Tools

6 tools
audit_mcp_tool_poisoning_and_eu_ai_actScan Remote MCP Server for Tool-Poisoning / Prompt-Injection & Classify EU AI Act RiskA
Read-onlyIdempotent
Inspect

Audits any external MCP tools/list schema for hidden prompt-injection, credential exfiltration, and Unicode steganography, plus 2026 EU AI Act Annex III compliance.

Inputs: mcpUrl, useCase.

audit_package_cve_and_cisa_kevAudit Software Package for OSV CVEs & CISA Exploited VulnerabilitiesB
Read-onlyIdempotent
Inspect

Audits an npm, PyPI, Go, Cargo, or Maven package version against OSV.dev and US CISA Known Exploited Vulnerabilities.

Inputs: name (required), version, ecosystem.

batch_sbom_cve_audit_x402Batch Audit Up to 15 SBOM Dependencies ($0.006 USDC via x402)A
Read-onlyIdempotent
Inspect

Audits up to 15 dependency specs in parallel against OSV.dev + CISA KEV ($0.006 USDC on Base via x402).

Inputs: packages (required).

optimize_compute_energy_grid_dispatchOptimize AI Compute & Robotics Fleet Dispatch Across Global Power GridsB
Read-onlyIdempotent
Inspect

Ranks 6 global datacenter and robotics charging grids (ERCOT, PJM, CAISO, Frankfurt, UK, Seoul) by live solar/wind surplus, cooling PUE, and power cost.

Inputs: hours, workloadKw.

search_patent_academic_prior_artSearch Global Patent & Scholarly Prior-Art WorksA
Read-onlyIdempotent
Inspect

Searches 250M+ technical works for patent novelty and prior-art assessment, returning reconstructed abstracts, DOIs, and citation counts.

Inputs: query (required).

track_biotech_clinical_trialsTrack Live ClinicalTrials.gov Biotech & MedTech PipelinesB
Read-onlyIdempotent
Inspect

Queries ClinicalTrials.gov API v2 for active Phase 1-3 trials, sponsors, NCT IDs, and completion dates.

Inputs: query (required).

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

    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables an agent or MCP client to scan MCP server configs and AI-agent tool manifests on demand, exposing tools that list detection rules and report hidden prompt injections, data-exfiltration attempts, embedded secrets, deceptive Unicode, unpinned packages and shell-launching tools. Each finding comes with a severity and location so risky tool descriptions can be caught before an attacker exploits them.
      57 PyPI
      1
      Apache 2.0
    • A
      license
      Not graded
      quality
      D
      maintenance
      Scans MCP servers for security vulnerabilities, prompt injection, and tool poisoning, providing risk scores and protection.
      4
      MIT
    Try in Browser

    Glama MCP Gateway

    Add one secure layer between your agents and this server.