Deepseek Harness、Openclaw知识图谱记忆插件。2026年4月受邀发布在清华大学讨论会。Knowledge Graph + Memory；Knowledge Graph Context Engine for OpenClaw — extracts structured triples from conversations, compresses context 75%, enables cross-session experience reuse

Persistent memory for agents. Corrections stick, context compounds, every session starts smarter.

Local-first AI agent runtime with sandboxed sessions, MCP tools, memory, credentials, audit/replay, and a built-in console. Run OpenAI, Anthropic, MiniMax, DeepSeek V4, and OpenAI-compatible models on your infrastructure.

Memory for coding agents — Claude Code, Codex, Cursor and 17 others. Indexes the sessions they already wrote to disk, including months from before you installed it, and recalls them in any of them. No LLM, no embeddings, one local Go binary.

Bridge Deepseek-harness into Claude Code for review, critique, delegation, and session import.

You say it. AutoCode ships it. 48 skills. Code to deployment in one session. I-Lang v5.0 judgment + secret-safe deploys. Free forever.

Second-model AI auto-review for DeepSeek Harness approval requests: a read-only reviewer subagent returns structured allow/deny verdicts with reasons, fail-closed by default, fully auditable from the session log (approval/asked -> autoReview/verdict -> approval/decided).

Run multiple Claude Code and Codex sessions in parallel — a browser terminal grid that shows which agent needs you. Local, tmux-backed.

Persistent long-term memory for Claude Code via MCP — captures coding decisions, bugfixes, and context across sessions. Hybrid FTS5 + TF-IDF search with episode batching. Single SQLite DB, no external services. Alternative to claude-mem with 600x lower cost.

Bounded, layered, approval-gated, auditable cross-session memory for DeepSeek Harness (capability seam: ctx.memory + SQLite provider + memory tool + frozen snapshot injection)

Auditable, token-gated DeepSeek Harness remote gateway: mobile QR access, per-device sessions, Host/Origin rewrite, settings/credentials/directory support.

🤫 Token-lean sessions at the harness level. An easy to grasp output style, output-shrinking hooks, and log compression cut both input and output tokens.

Session memory for Claude Code. Auto-captures every session into a plain Obsidian markdown vault. Zero config. MIT. macOS + Linux.

High-reliability Feishu/Lark bridge for DeepSeek Harness — QR one-click auth, multi-mode agents, card-based commands, zero-loss outbox, media in/out, session-log doctor, reusable DSH Web GUI

Pith is the hook that makes Claude Code sessions last 3x longer.

Run Codex and Cursor as background workers inside Claude Code. Reviews, tasks, and rescues with tracked jobs, live progress, and session handoffs.

Web cockpit that drives multiple real Claude Code TUI sessions in parallel, with Telegram control

Multi-session terminal workbench plugin for DeepSeek Harness

Memory for Claude Code that survives the session boundary — install the plugin into any repo: a hot cache injected every session under three hook-enforced caps, per-session handoffs, audit-driven promotion into knowledge and rules, plus agent orchestration and QA layers. Zero deps.

Workflow-continuity bridge for AI coding agents. Wrap Claude Code or Codex in a PTY and supervise, approve, and redirect the session from any device — async. The terminal companion for CodeAgent Mobile.

CLI engine for Monomind — an open-source MCP server that extends Claude Code with a codebase knowledge graph (tree-sitter + SQLite), persistent memory, multi-agent task coordination, and session hooks. Apache 2.0 licensed, fully local.

Graph memory for AI agents — decisions, context, and session history that survive across every conversation. Works with any LLM.

Claude Code harness adapter for TanStack AI — run Claude Code as a chat backend with local tool execution and stateful sessions.

Context window optimization for AI coding agents. Sandboxes tool output (98% reduction), persists session memory, and enforces routing across 17 platforms via MCP + hooks.