Multi-source threat-reputation MCP server (VirusTotal, URLhaus, urlscan, AbuseIPDB). The enrichment layer for the yara-detection-harness. The harness detects malicious patterns; the natural next question in an investigation is 'this rule fired -- is the indicator it surfaced actually known-bad?' This server wraps several reputation APIs as MCP tools an LLM agent can call mid-investigation, returning a normalized verdict for an indicator (a file hash, URL, IP, or domain) instead of each provider's full raw report.
Single tool 'lookup_indicator' has a clear verb-noun name and reasonable description (127 chars, within 10-1024 range). Input schema is present with type and description for the 'indicator' parameter. However, output schema is completely undocumented, the tool description does not specify what fields the response contains (e.g., does it return per-source verdicts, a consensus score, timestamps?). This forces LLMs to guess at response structure and breaks downstream tool chaining. The description lacks WHEN to use this tool vs alternatives, prerequisites (API key requirements), and what 'normalized verdict' means concretely. No error handling guidance is documented, what happens if an indicator is not found, or if all sources fail? Parameter description is adequate but could specify accepted formats more precisely (e.g., 'MD5/SHA-1/SHA-256 as hex strings, IPv4 in dotted-quad notation').
Look up an indicator (file hash, URL, IP, or domain) across multiple reputation sources (VirusTotal, URLhaus, urlscan, AbuseIPDB) and return a normalized verdict.
Output schema completely undocumented. Tool description does not specify response structure (fields, types, per-source vs consensus format). LLMs cannot plan downstream actions or extract required data.
Tool description lacks WHEN/WHY guidance. Does not explain when to call this vs other enrichment tools, what 'normalized verdict' means, or what prerequisites exist (API key requirements for each source).
No error handling guidance. Description does not explain what happens if indicator is not found, if a source fails, or if all sources are unavailable. LLM has no recovery path.
Parameter format constraints not specified. 'indicator' description should explicitly state accepted formats: 'MD5/SHA-1/SHA-256 as hex strings (32/40/64 chars), IPv4 in dotted-quad notation (e.g. 192.0.2.1), URLs with http/https scheme, domain names (e.g. example.com)'.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 57 | 2026-07-28+ | v2 |
No pagination or result-limiting guidance. If multiple indicators are looked up in sequence, no documentation of rate limits, caching behavior, or whether results are paginated.