Overview
Memory conduction is a category of AI agent memory management distinct from memory storage and memory retrieval. The term was introduced in March 2026 to describe a class of agent memory failures that had been widely experienced but not formally classified.
Conduction failures occur when an agent operationally destroys its own memory through normal, non-malicious tool use — overwriting files, overloading the context window, triggering destructive compression, or fabricating context after memory loss. In each case, the stored memories remain intact on disk. The failure occurs in the active pathway between storage and the agent's working context.
The distinction is significant because conduction failures are frequently misattributed to storage or retrieval problems, leading users to install storage-focused solutions that do not address the underlying cause.
Three-Layer Taxonomy
Agent memory can be classified into three functional layers, each addressing a distinct failure domain:
| Layer | Function | Addresses | Maturity |
|---|---|---|---|
| Storage | Where memories are created, indexed, and persisted | Memory never written to disk; memory written but not indexed | High |
| Conduction | Whether memories survive contact with the agent's own operations | Write corruption, context overload, compaction casualty, bootstrap truncation, confabulation | Low |
| Retrieval | How memories are located when needed | Irrelevant results; outdated information surfaced; inability to find existing memories | High |
The current ecosystem has mature storage solutions (lancedb, mem0, Cognee, QMD, Graphiti, Lossless Claw) and rapidly improving retrieval mechanisms. The conduction layer remains largely unaddressed as a formal category.
The full taxonomy, with community evidence, architectural analysis, and a reference implementation, is documented in the whitepaper.
Conduction Failure Modes
Five documented failure modes have been identified through community reports, GitHub issues, and independent technical analysis. In each mode, memory storage is functioning correctly — the failure occurs in the pathway between storage and working context.
- Write Corruption. The agent uses a file creation tool instead of an append operation, replacing curated memory content with an abbreviated summary. Documented in GitHub Issue #6877 and multiple community guides.
- Context Overload. The agent re-reads full bootstrap files on every message. As files grow, per-turn token usage increases until the context window reaches capacity, triggering emergency compression. Measured token usage in affected systems exceeds 260,000 per prompt.
- Compaction Casualty. Lossy context compression summarizes detailed decisions, verbal instructions, and operational constraints into generic descriptions. Critical directives are dropped from the summary without notification.
- Bootstrap Truncation. Platform-enforced character limits (20,000 per file in OpenClaw) silently truncate files from the bottom. Rules added most recently — typically those learned from operational failures — are removed first.
- Post-Compaction Confabulation. Following context loss, the agent generates plausible but fabricated context rather than acknowledging the loss. This mode is the most difficult to detect because the agent's output appears confident and specific.
Read the Full Research
The definitions, failure modes, and taxonomy summarized above are drawn from a comprehensive research paper that includes community evidence, architectural analysis of the OpenClaw memory system, the Memory Redirect pattern, and detailed documentation of each failure mode with root-cause analysis.
Whitepaper: Memory Conduction — The Missing Layer in Autonomous Agent Memory Architecture →
Map of Agentic Memory Flow
The following diagram illustrates how memory moves through the three-layer taxonomy — from storage through conduction into the agent's working context — and where each failure mode disrupts the pathway.
Resources
- Whitepaper: Memory Conduction Research Full research paper
- Map of Agentic Memory Flow Interactive diagnostic map