security
Security
Every MCP risk signal in one place — CVEs, tool safety, drift, naming, licenses. Heuristic: review signals, not verdicts.
45 CRITICAL
3626 HIGH
531 MEDIUM
1202 LOW
11 NONE
Risk: servers ranked by their composite exposure score — synthesised from CVEs, inferred permissions, drift, supply-chain and abandonment signals. Heuristic and banded; a high grade is a 'review this' signal, never a verdict.
A fix already ships upstream but the server still runs an older, vulnerable version — remediation lag, actionable today.
- @frontmcp/adapters HIGH CVE-2026-39885 1.4.1 → 2.3.0
- @frontmcp/sdk HIGH CVE-2026-39885 1.4.1 → 2.3.0
- 1 E EvoMap/evolver 100
- 2 E danny-avila/LibreChat 97
- 3 E samanhappy/mcphub 93
- 4 E modelcontextprotocol/inspector 85
- 5 E evalsone/mcp-bridge 82
- 6 E titan-agent 82
- 7 E askqai/claude-recall 80
- 8 E helius-labs/core-ai 80
- 9 D @claude-flow/cli 79
- 10 D figma-console-mcp 79
- 11 D HeliosNova/nova 77
- 12 D abrahamADSK/flame-mcp 77
- 13 D goklab/guardvibe 77
- 14 D heshengtao/comfyui_LLM_party 77
- 15 D flowing-abyss/obsidian-hybrid-search 76