ranxianglei/billion-context
Native Hermes hook plugin registering compress/decompress/acp_status tools that route model traffic through billion-context's self-spawned context-compression proxy.
Awesome Hermes Plugins › Context Engines
electricsheephq/lcm-x
LCM-X (plugin name hermes-lcm-x, context engine lcm-x) replaces Hermes Agent's built-in one-shot context compression with a lossless alternative. Where Hermes' default compressor prunes older tool results and asks an auxiliary model to summarize aging conversation into an active prompt, discarding the structured path back to the compacted detail, LCM-X persists every message in a plugin-local SQLite store with full-text-search metadata, compacts older context into depth-aware summary nodes, and condenses those summaries further into a hierarchical DAG as history accumulates. The active prompt is assembled from the system prompt, the highest-value summaries, and a protected recent tail, while a family of fifteen agent tools (lcm_grep, lcm_recall, lcm_query_state, lcm_expand, lcm_status, lcm_doctor and others) lets the agent search, inspect and expand compacted material on demand rather than flooding the live context. Beyond the core compaction loop, three opt-in feature families extend it into a fuller memory system: large-output externalization that moves oversized tool, media or raw payloads to recoverable references; temporal memory with day/week/month rollups and natural-time recall; and semantic retrieval with embedding-backed hybrid search against Voyage, OpenAI-compatible endpoints, Ollama, or an in-process FastEmbed model. Two independent privacy layers are available: opt-in durable redaction of secrets such as API keys and private keys before anything is stored, and automatic protection of the copies sent to cloud embedding providers regardless of that setting. The project is based on a published paper on lossless context memory and draws on a comparable OpenClaw project, lossless-claw, for which it ships an import path. It targets Hermes users running long-lived sessions who need exact recall of earlier decisions and tool output after compaction, and ships diagnostics, backup-first repair tooling and optional slash commands for day-to-day operation.
https://github.com/electricsheephq/lcm-x
Also in Context Engines
stephenschoettler/hermes-lcm
hermes-lcm is a lossless context management plugin for Hermes Agent, providing a DAG-based context engine that ensures no messages are lost through persistent storage and intelligent summarization.
claudioemmanuel/squeez
squeez is an end-to-end token optimizer that provides hook-based token compression for Hermes Agent and several other AI CLI hosts, increasing context window efficiency.
EfficientContext/ContextPilot
ContextPilot is a context engine that accelerates long-context LLM inference by optimizing context reuse, offering significant speedups and token savings for RAG, memory chat, and agentic AI worklo...
beniedev/moonbite
Experimental Hermes Agent plugin adding cross-session memory, short-lived working state, and bounded, verified autonomous action on top of the host agent.
Elevatormusic/hermes-warm-compaction
Context-engine plugin for Hermes Agent that speeds up compaction by writing the handoff summary against the cached prompt prefix instead of the whole transcript.
byenzyme/enzyme-skill
A Hermes plugin that compiles a Markdown vault into a local concept graph and injects trending entities and semantic search results into the session through three lifecycle hooks and five tools.
eas4ai/hermes-context-manager
Silent-first context optimization plugin for the Hermes Agent gateway that compresses tool outputs, deduplicates repeated work and summarizes completed phases without the main model noticing.