Show HN: Hermes-agentmemory, pull-model episodic memory with real deletes (github.com)

The Hermes Agent has introduced a new plugin called "agentmemory," which enhances its episodic memory capabilities by implementing a pull-model that prioritizes real deletion of memory items. Unlike traditional memory backends that consolidate episodes into summaries, potentially retaining deleted information, the agentmemory plugin ensures immediate and complete deletion of memories, offering users an audit trail of what events were used in prompt generation. This shift addresses key limitations in existing memory systems where recall can become opaque and unreliable. This development is significant for the AI/ML community as it allows for a more transparent and accountable memory management system within AI agents. By enabling users to audit memory entries and ensuring deletions are effective, the plugin enhances the reliability of AI-generated responses and supports ethical considerations around memory retention. The plugin operates synchronously, without background processes, leading to a slight initial delay but offering a clear trade-off: real deletions, consistent summarization quality, and a straightforward coding implementation that can be easily integrated. This innovation positions Hermes as a more flexible and responsible self-hosted AI solution, catering to users' needs for memory management while fostering growth and adaptability.