All-in-one workflow plugin—loops, swarms, and teams on Claude Code's Task System. All enforce exit criteria—swarm is faster with parallel queue execution, teams add contract-first coordination. Plan your way, execute your way. Optional: Beads for persistent memory, Ralph TUI for dashboard.

When to delegate to `cavecrew-investigator` (locate code), `cavecrew-builder` (1-2 file edit) or `cavecrew-reviewer` (diff review) instead of working inline or using `Explore`. Their output is compressed, so main context lasts longer.

Strategic BDI reasoning engine for autonomous AI agents — goal decomposition, utility scoring, risk evaluation, and reactive replanning over PuterVision memory and world models.

Agentic Version Control System — an AI-native VCS where intent, session, operation, evidence and decision are first-class objects and code is a materialized projection of an operation graph.

Stop your AI from making things up — it proposes, deterministic tools decide, every claim checked against ground truth with evidence. Grounded facts and context survive resets. Reverse engineering is the proving ground. MCP server + CLI.

Harness the power of local LLMs with this TUI MCP Client for Ollama. Featuring all core MCP primitives (tools, prompts, resources), agent mode, multi-server, model switching, streaming responses, human-in-the-loop, thinking mode, model params config, system prompts, and saved preferences.

Strategic BDI reasoning engine for autonomous AI agents — goal decomposition, utility scoring, risk evaluation, and reactive replanning over PuterVision memory and world models.

Compress a memory file such as CLAUDE.md or a todo list into caveman format to save input tokens, keeping a readable backup. Sends the file's text to the configured model provider. Trigger: /caveman-compress.

Show recorded output and cache-read token usage and mode attribution for the current Claude Code session, or locate the host's native usage report. Trigger: /caveman-stats.

Act on a Caveman learn report - review the ranked token sinks, apply cost-lowering fixes with per-edit consent, and report what those fixes returned. Use when asked to lower an agent's token cost, what caveman has saved, to trim a heavy CLAUDE.md, or to offload re-pasted context into cavemem.

Read-only repository explorer for cold-start orientation, broad cross-file localization, or when a direct search failed. Skip it when the exact file or symbol is already named. Returns path:line citations only; its reads stay out of main context.

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

Enables AI coding agents across supported harnesses to run a structured multi-agent workflow with persistent memory, squad roles, and bundled skills through MCP tools.

Enables AI coding agents to extract structural C# contracts from files and build solution-wide dependency graphs, so they can navigate code and assess change impact without reading entire files.

Agent plugin for major harnesses — DeepSeek Harness (DSH), OpenCode, Pi and Codex: define custom AI agent roles with per-role prompts, models, skills, permissions, memory and multi-agent orchestration.

Builds up to a goal in a field you have no expertise in: splits it into the foundations that goal actually needs, mobilizes what you already hold, learns what is missing, and leaves every solved path behind for the next goal. Nothing is called done until a check passes.

终结 Agent 无限规划：认知循环让判断有来处、反馈有归因、规划必收口；记忆完全本体化（v0.4.4+）——写侧三本体条目（event 记忆本体 / fact 知识本体 / rule 规则本体）+ ontology 投影一等公民，注入段按序承载（知识→规则→事件→冷储）；判分身份职责入架构（口径一致、判据绑定来源、同题同标，漂移清单沉入判分领域本体）；联动 linghun-assembler A2A 记忆管理团队，让子智能体的判断履历反哺主脑。不是人设卡，是判断力——给 DeepSeek Harness

Git status badges for DeepSeek Harness: an input-row chip and a sidebar session-row action glyph, on every install — no patch and no manual step. Event-driven freshness over SSE.

Rolling context compression for Claude Code — never hit the context wall. Auto-compresses old messages while keeping recent context verbatim. Zero config, zero latency. Works as a Claude Code plugin.

Model Context Protocol server transport: serve typed procedures as MCP tools to Claude and other MCP clients, with streaming progress, list_changed and an expose policy.

Enables AI agents to debug Commodore software in a locally running VICE emulator through its text remote monitor, inspecting registers, memory and screens, disassembling code, setting breakpoints and watchpoints, controlling execution, and working with disk images, snapshots and exported graphics. Runs locally over stdio and includes safety guidance for state-changing operations.

Enables AI clients to drive local video editing projects through a CEP bridge, covering media ingest, sequence and timeline editing, effects, keyframes, captions, markers, audio, color, and exports across hundreds of discoverable tools. Ships context resources, guided prompts, and higher-level assembly workflows for agent-driven editing.

Context compression for AI coding agents: compresses tool output before it enters the model, dedups repeats to 13-token refs. Claude Code, Cursor, Codex, Kiro, Zed, any MCP client. Rust, zero LLM calls.

MCP server for Roblox Studio. 35 tools, push-based SSE bridge, two-way script sync with files, editor-safe script edits, one-step undo.