hybrid-memory

name	hybrid-memory
description	Recalls past conversations, retrieves stored facts, and searches documents using OpenClaw QMD vector memory and Graphiti temporal knowledge graph. Use when the user asks to remember something, recall a previous discussion, look up what was said before, find prior context, check conversation history, or retrieve a stored fact or document.

Hybrid Memory System

Combines two memory backends into a single recall layer:

1. QMD (Vector Store): Retrieves documents, specs, and full-text content by semantic similarity.
2. Graphiti (Knowledge Graph): Retrieves temporal facts, entity relationships, and cross-agent knowledge.

Primary Tool

For most memory queries, use the hybrid search script. It queries both systems in parallel and merges results.

~/.openclaw/scripts/memory-hybrid-search.sh "your query"

Optional flags:

• [group_id] — Specify agent group (default: openclaw-main)
• --json — Output JSON for programmatic use

Specific Tools (Advanced)

Use these when the hybrid script returns no results, or when you need granular control over a single backend.

Graphiti Only (Temporal/Facts)

Search for specific temporal facts:

~/.openclaw/scripts/graphiti-search.sh "your question" openclaw-main 10

Log new facts (IMPORTANT — do this after key conversations or decisions):

~/.openclaw/scripts/graphiti-log.sh openclaw-main user "Name" "Fact to remember"

QMD Only (Deep Document Search)

If you need more results or specific file filtering:

qmd search "query" -n 10

Recall Pattern

1. User asks question ("What was the plan for the project?")
2. Run Hybrid Search (~/.openclaw/scripts/memory-hybrid-search.sh "plan for the project")
3. Synthesize Answer from both the temporal facts and document snippets found.
4. If needed: Use read to get the full content of a file found in the QMD results.
5. If no results: Retry with broader terms, try Graphiti direct (graphiti-search.sh), or widen QMD search (qmd search "query" -n 20).

Proactive Logging

Keep the knowledge graph current by logging important facts after they occur:

• After the user shares a preference, decision, or key fact
• After completing a task with outcomes worth remembering
• After learning new entity relationships (people, projects, systems)

~/clawd/scripts/graphiti-log.sh clawdbot-main user "UserName" "User decided to use PostgreSQL for the new service"

When to Use Which

Question Type	Use
"What's in GOALS.md?"	Hybrid search → read file
"When did we discuss X?"	Hybrid search (Graphiti results)
"What did I say last Tuesday?"	Graphiti direct
"Find notes about architecture"	Hybrid search (QMD results)
"Remember that I prefer dark mode"	graphiti-log.sh to store the fact