ai-smithery-alex-llm-attack-mcp-server
ToolProof Connection Signal
Review First
Do not connect casually. Review the exposed tools, credentials, network/file/API access, and approval path before giving an agent access.
Evidence Confidence86/100
Tools Inferred11
Control Signals0
Profile status
ToolProof reviewSnapshotEvidence-backed review from observed public information.
Owner statusUnclaimedOwner-provided details have not been maintained yet.
Watch statusNot monitoredEnable ToolProof Watch to monitor drift and freshness.
Last observed2026-06-07T19:20:29.201592+00:00MCP servers can change after observation.
Why this signal
- Source evidence fetched from 3 file(s).
- 11 MCP tool(s) inferred: create_http_app, get_all_tactics, list_tactics, query_attack_technique, query_detections.
- No visible control signals were observed.
- 3 static finding(s) should be reviewed before connection.
What this server exposes
create_http_appget_all_tacticslist_tacticsquery_attack_techniquequery_detectionsquery_mitigationsquery_techniqueread_commit_hashresolve_log_levelsearch_technique_fullserver_info
Source and observation
Source: https://github.com/alex-llm/attAck-mcp-server
Observed: 2026-06-07T19:20:29.201592+00:00
Machine ID: mcp:ai-smithery-alex-llm-attack-mcp-server
Capabilities inferred
- destructive_capability
- elevated
- {"has_packages": false, "has_remotes": true, "package_registry_types": [], "remote_types": ["streamable-http"], "requires_secret_header": false}
Controls observed
- No controls observed.
Recommended controls
- Install in a sandbox before team or production use.
- Pin the exact package/repository version.
- Review install scripts, Dockerfile behavior, and dependency pins.
Open questions before connection
- What credentials or environment variables are required?
- What data can leave the local environment?
- What systems can the agent reach through this server?
- What actions should require human approval?
Findings and concerns
- Install risk pattern: unpinned_dependency — Dockerfile:7
- Detected capability: filesystem_read — main.py:51
- Detected capability: network_egress — main.py:875
Claim or update this profile
Are you the owner of this MCP server? Submit official docs, intended use cases, support links, security contact, and recommended safe configuration.
Owner-provided information is shown separately from ToolProof evidence and does not change the connection signal unless reviewed.
Payment may buy review, monitoring, or sponsored visibility. It does not change the ToolProof connection signal.
Evidence fetched
- https://github.com/alex-llm/attAck-mcp-server
- https://github.com/alex-llm/attack-mcp-server
- Registry entry: ai.smithery/alex-llm-attack-mcp-server 2.1