RSS & API
Everything on ThreatDeck is available to your tools and agents. Public, read-only, no key required. Please cache responses (they change about every 15 minutes). For integrations, get a free API key for higher limits.
Feeds
| /rss.xml | RSS 2.0, latest briefs |
| /feed.json | JSON Feed 1.1, latest briefs |
REST
/api/v1/stories?tab=top|latest|exploited|ai|breaches|research&q=&since=&limit= | Story feed (one-liners, actions, entities, IOCs) |
/api/v1/dashboard | Panels: critical CVEs, KEV, patch priorities, ransomware aggregates, trends |
/api/v1/cards | Deep briefs with verified citations |
/api/v1/deck/latest | Today's deck |
/api/v1/tools/:slug/check?version=x | Is this AI dev tool version safe? |
/api/v1/iocs?format=json|csv|stix | Indicators of compromise (STIX 2.1 bundle available) |
/api/v1/search?q= | Search stories, briefs and CVEs |
/api/v1/techniques/:id | OWASP LLM / Agentic and MITRE ATLAS technique with stories, tools, CVEs |
/api/v1/radar?side=&kind=&q= | Agentic security tool radar (defender notes for offensive tools) |
/api/v1/mcp-servers/lookup?pkg=npm:<name>@<version> | Is this MCP server package affected by an advisory? (also ?repo=, ?url=, ?name=) |
/api/v1/detections?technique=&language= | Hunting queries (Sigma, KQL, SPL, ES|QL, CloudTrail SQL, osquery, YARA) |
/api/v1/playbooks/:slug | Playbook: patterns, checklist, stack, vendor POC questions |
/api/v1/ai-found-vulns?system=&since= | Vulnerabilities credited to AI systems, with evidence |
/api/v1/feed.xml?follow=tool-<slug>,technique-<id>,<vendor> | Personal RSS for what you follow |
MCP
Streamable HTTP endpoint: https://threatdeck.net/mcp. Tools include get_top_stories, get_dashboard_panels, search, get_iocs, get_latest_deck, check_ai_tool_version, check_mcp_server, get_technique, search_tools, get_detections and get_playbook. Ask your agent “is the MCP server I'm about to install affected by an advisory?” or “give me KQL to hunt for LLMjacking”. See methodology for how data is produced.
Get an API key
Optional and free. Keys raise the rate limit on /api/v1/* and /mcp; no account or email needed.
| Without a key | 60 requests per minute per IP address |
| With a key | 600 requests per minute per key |
| Over the limit | HTTP 429 with retry-after: 60 and x-ratelimit-limit. Limits are counted per Cloudflare location and are approximate. |
| Bad or revoked key | HTTP 401 (never a silent fallback to keyless limits) |
Your API key
Copy it now: it is shown once. ThreatDeck stores only a hash and this browser does not keep it.
Use it
Send it as Authorization: Bearer tdk_… (or x-api-key: tdk_…).
curl -H "Authorization: Bearer tdk_…" "https://threatdeck.net/api/v1/stories?tab=ai&limit=5"
MCP clients that support headers (Streamable HTTP):
{
"mcpServers": {
"threatdeck": {
"type": "http",
"url": "https://threatdeck.net/mcp",
"headers": {
"Authorization": "Bearer tdk_…"
}
}
}
} Claude Code:
claude mcp add --transport http threatdeck https://threatdeck.net/mcp --header "Authorization: Bearer tdk_…"