Search & browse
Semantic search and entity browsing from the CLI.
These commands query the local knowledge graph. Your AI agent uses MCP tools
(search_code, search_document, etc.) for the same data during sessions.
For code, MCP search_code returns graph-enriched source and optional inline
annotations — see Knowledge layer.
orkai search
Hybrid semantic + keyword search across indexed entities.
orkai search "how does authentication work"
orkai search "connection pool" --type=code --lang=go
orkai search "rate limiting" --brief --limit=5
orkai search "auth" --global --json Flags
| Flag | Description |
|---|---|
--type types | Filter by entity type (comma-separated) |
--lang langs | Filter by language |
--category id | Search within one project category |
--global | Search all categories |
--limit n | Max results (default 10) |
--brief | One line per result |
--json | Raw JSON response |
--graph-hops n | Graph expansion depth (default 1) |
orkai get
Fetch one entity by ID (full or partial ID).
orkai get a1b2c3d4 orkai list
Browse entities with filters, or list categories and orphans.
orkai list
orkai list --type=code --lang=go
orkai list --type=document --limit=5
orkai list --categories
orkai list --orphans --count
orkai list --global --count Flags
| Flag | Description |
|---|---|
--categories | List projects (categories) instead of entities |
--orphans | Entities with no category |
--count | Count only, no rows |
--json | Raw JSON |