explore

/explore

You are exploring your organization's feedback taxonomy and data — the hierarchical classification system for customer feedback. This is a utility command that outputs directly in chat (no report generation).

Pre-Flight

1. Check if context/organization.json exists. If not, tell the user: "Run /start first to connect to your organization's Knowledge Graph." Stop.
2. Call get_organization_details from the enterpret-wisdom-mcp MCP server.
3. If it fails with an auth error, tell the user to run /start and stop.
4. If successful, read context/organization.json for org name and L1 categories.

Skills (reference if needed, do NOT read upfront)

• wisdom-kg — query patterns and KG rules (only read if a query fails and you need to debug)

Behavior

No argument: Show L1 overview

If the user runs /explore with no argument:

1. If context/organization.json already has L1 categories, present those directly. Otherwise query:

MATCH (nli:NaturalLanguageInteraction)-[:SUMMARIZED_BY]->(fi:FeedbackInsight)-[:HAS_TAGS]->(cft:CustomerFeedbackTags)-[:BELONGS_TO_L1]->(l1:L1)
RETURN l1.name AS category, COUNT(DISTINCT fi.feedback_record_id) AS volume
ORDER BY volume DESC
LIMIT 20

1. Present as a numbered table:

#	Category	Volume
1	{top category}	{volume}
2	{second category}	{volume}
...	...	...

1. Say: "Pick a category number or name to drill deeper, or say 'done' to exit."

With L1 argument: Show L2 breakdown

If the user runs /explore {category} or selects a category:

1. Query L2 under that L1:

MATCH (fi:FeedbackInsight)-[:HAS_TAGS]->(cft:CustomerFeedbackTags)-[:BELONGS_TO_L1]->(l1:L1)
MATCH (cft)-[:BELONGS_TO_L2]->(l2:L2)
WHERE l1.name = "{L1_CATEGORY}"
RETURN l2.name AS subcategory, COUNT(DISTINCT fi.feedback_record_id) AS volume
ORDER BY volume DESC
LIMIT 20

1. Present as tree + table:

{L1 Category}
├── {L2 subcategory 1} ({volume})
├── {L2 subcategory 2} ({volume})
├── {L2 subcategory 3} ({volume})
└── ...