Giving AI Agents a Memory of Your Crisp Inbox

Description and use

AI agents can now remember Crisp inbox conversations via a Cognee connector that aggregates threads, uses full‑snapshot sync, and supports incremental updates.

Problem

Chat inboxes like Crisp produce short, numerous sessions without delete feeds, making it hard to build a persistent knowledge graph for AI agents.

Outcome

Agents can now answer questions about Crisp inbox content, including summarizing billing complaints, thanks to a connector that aggregates conversations, performs full‑snapshot sync, and handles incremental updates.

Workflow

• 1. Install connector: pip install cognee-community-connector-crisp
• 2. Set environment variables CRISP_IDENTIFIER, CRISP_KEY, CRISP_WEBSITE_ID
• 3. Import cognee and crisp_source
• 4. Call cognee.remember(crisp_source(), dataset_name="crisp", max_rows_per_table=0)
• 5. For incremental sync, pass since= watermark
• 6. Query memory: cognee.search(query_text="What did customers complain about this month?", query_type=cognee.SearchType.GRAPH_COMPLETION, datasets=["crisp"])

Evidence

Cognee has a "forget-on-delete" mechanism: if a record drops out of the snapshot between two syncs, its orphan-cleanup removes it from the graph.

Source

The connector aggregates an entire conversation into a single markdown document — visitor context plus the full transcript:

Source

Cognee already has a "forget-on-delete" mechanism: if a record drops out of the snapshot between two syncs, its orphan-cleanup removes it from the graph.

Source