AgentDB Query

Query and manage AgentDB through the controller bridge. AgentDB exposes 20 agentdb_* MCP tools; this skill enumerates the standard usage path.

When to use

When you need to store, retrieve, or search knowledge across agent sessions. AgentDB provides hierarchical storage, causal knowledge graphs, semantic routing, and context synthesis.

Steps

1. Check health — mcp__plugin_ruflo-core_ruflo__agentdb_health. Sanity-check available: true.
2. Start session — mcp__plugin_ruflo-core_ruflo__agentdb_session-start if not already active.
3. Store knowledge — mcp__plugin_ruflo-core_ruflo__agentdb_hierarchical-store for structured tier-keyed data (tiers: working|episodic|semantic).
4. Recall knowledge — mcp__plugin_ruflo-core_ruflo__agentdb_hierarchical-recall with a query.
5. Search patterns — mcp__plugin_ruflo-core_ruflo__agentdb_pattern-search for learned patterns (ReasoningBank-routed).
6. Summarize past task outcomes — mcp__plugin_ruflo-core_ruflo__agentdb_context-synthesize for the success rate, reward and critique patterns of earlier tasks like this one. Not for combining ordinary memories: it summarizes validated task episodes (task, reward, success) and answers no-memories or no-eligible-episodes for ordinary records, which is correct, not missing data. Combine those yourself from what recall and search return.
7. Build causal graph — mcp__plugin_ruflo-core_ruflo__agentdb_causal-edge to link related knowledge.

Available controller groups

Call mcp__plugin_ruflo-core_ruflo__agentdb_controllers to list the runtime registry. Functional categories surfaced via the 20 MCP tools:

• Hierarchical — agentdb_hierarchical-store, _recall (tier-routed)
• Pattern — agentdb_pattern-store, _search (ReasoningBank-routed)
• Semantic — agentdb_semantic-route, _context-synthesize
• Causal — agentdb_causal-edge (graph-node backend with bridge fallback)
• Lifecycle — agentdb_health, _controllers, _session-start, _session-end
• Bulk — agentdb_batch (≤500 entries), _consolidate
• Quality — agentdb_feedback
• Graph reads — agentdb_graph-query (k-hop / semantic / pagerank from a domain-prefixed node id such as entity:xyz; SQL k-hop is bounded at 3 hops, check appliedDepth and truncated) and agentdb_graph-pathfinder (personalized-pagerank by default; dynamic-mincut returns the seed's side of an exact min cut with cutValue, cutEdges and backend; see the graph-mincut skill)
• Deletes (destructive) — agentdb_hierarchical-delete, agentdb_causal-edge-delete, agentdb_causal-node-delete. Confirm the exact ids with the user before calling; they are not undoable from this plugin.

Important: namespace handling

Namespace strings apply to memory_* and embeddings_search only. The agentdb_hierarchical-*, agentdb_pattern-*, and agentdb_causal-edge tools route by tier or controller, not namespace. Don't pass namespace: 'foo' to those tools — it will be silently ignored. See plugin README "Namespace convention".

Operational fallbacks (branch on these)

• controller: 'memory-store-fallback' — pattern persisted via memory_store --namespace pattern. NOT a failure.
• _graphNodeBackend: true — causal-edge handled by @ruvector/graph-node.
• success: false, error: '...Use memory_store/memory_search instead.' — bridge unavailable; switch to memory_* tools per the README replacement table.

CLI alternative

npx @claude-flow/cli@latest memory search --query "your query" --namespace patterns
npx @claude-flow/cli@latest memory store --key "key" --value "value" --namespace patterns
npx @claude-flow/cli@latest memory list --namespace patterns