Skip to main content

ACL-aware search

Search is the fastest way into your team’s knowledge. Premora combines keyword and vector retrieval, and every query is filtered so it only ever returns what the asking user is permitted to see.

How it works​

  • Keyword + vector — results blend an OpenSearch keyword index with a Qdrant vector index, so you can search by exact terms or by meaning.
  • ACL filtering — results are filtered against cached entitlement snapshots captured from the source systems at ingest. If you can’t see something in Confluence, Jira, or Git, you won’t see it in Premora search either. Retrieval fails closed.
  • Citations & lineage — every result links back to its source. You can see where a fact came from and how fresh it is before you trust it.
  1. Enter a natural-language question or keywords.
  2. Scan the ranked results — wiki pages and chunks, each with its source and freshness.
  3. Open a result to read the full page, then follow its lineage to the originating source or its graph neighborhood to discover adjacent topics.

Why it’s trustworthy​

Premora is the tool that doesn’t guess. Rather than synthesizing a confident answer with no provenance, search leads with the source. When you need synthesis across multiple sources, use the analyst workbench; when a coding agent needs the same grounded context, it uses the agent skill.