MISP MCP provides 11 read-only tools with consistent naming (verb_noun pattern: search_*, list_*, get_*) and reasonable descriptions (avg ~120 chars). All tools have input schemas with typed parameters and descriptions. However, output schemas are not documented, LLMs cannot predict response structure for downstream chaining. Parameter descriptions lack format constraints (e.g., date_from/date_to state 'YYYY-MM-DD' but no regex/pattern enforcement visible). Error handling is absent, no recovery guidance, no actionable error messages. No tool annotations (readOnlyHint present conceptually but not declared in schema). Composition is sound: tools are single-responsibility and chainable via IDs. Security is adequate for read-only: no secrets in params, but no audit logging visible.
Get a specific galaxy with its clusters. Returns the galaxy definition and optionally all its clusters (e.g. all threat actors in the threat-actor galaxy).
Get a specific taxonomy with all its entries. Returns the full taxonomy including all predicates and values that can be used as tags.
List all available MISP taxonomies. Taxonomies provide standardised tag vocabularies for classifying threat intelligence (e.g. TLP, admiralty scale, kill chain, CIRCL taxonomy).
Search MISP attributes directly. Returns individual attributes (indicators) across all events. More granular than event search when looking for specific IOCs.
Search the MISP event index (lightweight). Returns event metadata without full attribute/object data. Faster than search_events for browsing and filtering.
Output schemas not documented. LLMs cannot predict response structure (fields, types, nested objects) for downstream tool chaining. E.g., search_events returns dict but structure is opaque.
No error handling or recovery guidance. Tools lack actionable error messages (e.g., 'API key invalid' vs 'Authentication failed. Check MISP_API_KEY env var'). No categorization of retryable vs fatal errors.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 59 | 2026-07-28+ | v2 |
Search MISP events. Use this to find threat intelligence events matching specific criteria. Returns event data including attributes and objects.
Search MISP feeds. Feeds are external sources of threat intelligence that MISP can pull data from (OSINT feeds, commercial feeds, other MISP instances).
Search MISP galaxies by name or description. Galaxies are knowledge bases of threat actors, malware, attack techniques (MITRE ATT&CK), tools, and other threat intelligence concepts.
Search clusters within a specific galaxy. Clusters are individual entries in a galaxy, e.g. specific threat actors, malware families, or ATT&CK techniques.
Search MISP objects. Objects group related attributes together (e.g. a file object contains filename, hash, size).
Search MISP tags by name. Tags are used to classify and filter events and attributes. Common tag formats include taxonomies (tlp:white), galaxies (misp-galaxy:threat-actor= "APT28"), and custom tags.
Parameter format constraints not enforced. date_from/date_to describe 'YYYY-MM-DD' but no regex pattern or validation visible. LLMs may pass invalid dates (e.g., '2024-13-45'). limit/page parameters lack min/max bounds.
No tool annotations (readOnlyHint, idempotentHint). All tools are read-only but this is not declared in schema. Agents cannot distinguish safe-to-retry tools from stateful ones.
No pagination guidance in descriptions. search_events/search_attributes accept limit/page but descriptions do not explain default behavior, max result size, or when pagination is required.