Proactive Surfacing

Traditional RAG systems only provide relevant information when the agent explicitly requests a search. memtomem-stm’s proactive surfacing observes every tool call the agent makes and automatically searches for and injects relevant memories into responses.

How It Works

When an agent calls an MCP tool, the STM proxy runs this pipeline:

Tool call → Context extraction → LTM search → Relevance gating → Inject into response

No agent code changes needed — just routing through the STM proxy enables automatic memory injection for all MCP communication.

5-Level Context Extraction

How search queries are extracted from tool calls, by priority:

Priority	Method	Description
1	Tool-specific query template	Pre-defined query patterns mapped to tool names
2	_context_query argument	Explicit search query passed by the agent
3	Path arguments	Context extracted from file paths, URLs, etc.
4	Semantic keys	Semantic keyword combination from tool arguments
5	Tool name	Last resort — use the tool name itself as query

Relevance Gating

Surfaced memories are filtered through 5 stages to ensure usefulness:

1. Context extraction — Can a meaningful query be generated from the tool call?
2. Relevance assessment — Is the extracted query suitable for memory search?
3. LTM search — Hybrid search for candidate memories
4. Score filtering — Remove results below the min_score threshold
5. Deduplication — In-session + cross-session (7-day) duplicate prevention

Model-Aware Defaults

Automatically scales based on the agent’s context window size:

Context window	Compression	Injection size	Result count
≤ 32K	High compression	Small	Few
32K – 200K	Default	Medium	Default
> 200K	Low compression	Large	Many

Feedback Loop

When an agent evaluates surfacing quality, the auto-tuner continuously optimizes per-tool relevance thresholds:

• helpful → Maintain or lower min_score for that tool
• not_relevant → Raise min_score (stricter filtering)
• already_known → Expand deduplication window

Safety Mechanisms

• Circuit breaker (3-state) — Temporarily suspend surfacing after consecutive failures
• Write-tool skip — Disable surfacing for tools with side effects (file writes, deletes)
• Query cooldown — Prevent rapid repeated searches of the same query
• Sensitive data detection — Auto-detect and filter API keys, passwords, PII