Dynamic multi-tenant OpenAPI adapter for providing context to AI models
38 tools with minimal descriptions (all ~20 chars or less) and no visible input parameter descriptions. All tools are READ_ONLY with empty required arrays, suggesting auto-generated from OpenAPI spec without LLM-specific optimization. No error handling guidance, no output schemas documented, no parameter constraints (enums, ranges). Tool names follow verb_noun pattern (positive), but descriptions lack context for LLM selection. No evidence of composition patterns or chaining support.
Get all channels
Get all connections
Get all content stores
Get all dialogues
Get all group types
Get all groups of a given group type
Get all lifecycles
All 38 tool descriptions are under 20 characters ('Get X' or 'Get all X'). Descriptions lack WHEN to use, prerequisites, or context for LLM selection.
No input parameter descriptions visible in schema. Parameters like 'contentStore', 'connection', 'dialogue', 'profileId' have type and description in JSON, but LLM cannot infer whether these accept IDs, names, emails, or slugs.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 39 | <=2025-11-25 | v2 |
Get all listeners
Get all models
Get all objectives
Get all plugins
Get all profile or group properties
Get all timeline event rollups
Get all segments
Get audit events
Get runs for a specific connection
Get content items from a specific content store
Get dialogue statistics
Get interactions
Get ONNX binary for a specific model
Get a specific channel
Get a specific connection
Get a specific dialogue
Get a specific group of a given group type
Get a specific lifecycle
Get a specific listener
Get metadata for a specific model
Get a specific objective
Get a specific plugin
Get a specific profile
Get a specific profile or group property
Get a specific timeline event rollup
Get a specific timeline event type
Get events for a specific profile
Get profiles in a specific segment
Get recommendations
Get all timeline event types
Search for profiles
No output schemas documented. Tools return data but LLMs cannot plan downstream calls or extract fields. LLMs need to know what fields to expect.'
No pagination parameters (limit, offset, page, cursor) visible in any tool. Tools like getAllConnections, getAllDialogues, getAllPlugins will return unbounded results, risking context window exhaustion.
No error handling guidance. Tools do not document what happens on 404, 401, timeout, or invalid input.