# CrabTalk Documentation ## Daemon - [Installation](https://crabtalk.ai/docs/crabtalk/installation) - [Quickstart](https://crabtalk.ai/docs/crabtalk/quickstart) - [Tutorials](https://crabtalk.ai/docs/crabtalk/tutorials) - [Configuration](https://crabtalk.ai/docs/crabtalk/config) - [REPL](https://crabtalk.ai/docs/crabtalk/repl) - [Hub](https://crabtalk.ai/docs/crabtalk/hub) - [Auth](https://crabtalk.ai/docs/crabtalk/auth) - [Providers](https://crabtalk.ai/docs/crabtalk/providers) - [Commands](https://crabtalk.ai/docs/crabtalk/commands) - [Agents](https://crabtalk.ai/docs/crabtalk/agents) - [Sessions](https://crabtalk.ai/docs/crabtalk/sessions) - [Cron](https://crabtalk.ai/docs/crabtalk/cron) - [Events](https://crabtalk.ai/docs/crabtalk/events) - [Skills](https://crabtalk.ai/docs/crabtalk/skills) - [Memory](https://crabtalk.ai/docs/crabtalk/memory) - [MCP Servers](https://crabtalk.ai/docs/crabtalk/mcp-servers) - [Protocol](https://crabtalk.ai/docs/crabtalk/protocol) ## CrabLLM - [Introduction](https://crabtalk.ai/docs/crabllm/introduction) - [Getting Started](https://crabtalk.ai/docs/crabllm/getting-started) - [Configuration](https://crabtalk.ai/docs/crabllm/configuration) - [Docker](https://crabtalk.ai/docs/crabllm/docker) - [Providers](https://crabtalk.ai/docs/crabllm/providers/overview) - [OpenAI](https://crabtalk.ai/docs/crabllm/providers/openai) - [Anthropic](https://crabtalk.ai/docs/crabllm/providers/anthropic) - [Google Gemini](https://crabtalk.ai/docs/crabllm/providers/google) - [Azure OpenAI](https://crabtalk.ai/docs/crabllm/providers/azure) - [AWS Bedrock](https://crabtalk.ai/docs/crabllm/providers/bedrock) - [Ollama](https://crabtalk.ai/docs/crabllm/providers/ollama) - [Routing](https://crabtalk.ai/docs/crabllm/features/routing) - [Streaming](https://crabtalk.ai/docs/crabllm/features/streaming) - [Authentication](https://crabtalk.ai/docs/crabllm/features/auth) - [Extensions](https://crabtalk.ai/docs/crabllm/features/extensions) - [Storage](https://crabtalk.ai/docs/crabllm/features/storage) - [Management](https://crabtalk.ai/docs/crabllm/features/management) - [Architecture](https://crabtalk.ai/docs/crabllm/architecture) - [Libraries](https://crabtalk.ai/docs/crabllm/libraries) ## Blog - [L0 to L5: where AI agents land on the autonomy spectrum](https://crabtalk.ai/blog/autonomy-spectrum): A survey of how AI agents are shifting from human-approved to human-supervised — autonomy levels, product positioning, and what it means for agent runtimes. - [MiroFish: swarm prediction through social simulation](https://crabtalk.ai/blog/mirofish-swarm-prediction): We examined MiroFish's architecture — from OASIS swarms to GraphRAG extraction. How it predicts the future and where the approach breaks down. - [Async compaction: the race conditions nobody talks about](https://crabtalk.ai/blog/async-compaction-challenges): We examined how frameworks handle background compaction. Most avoid it entirely. The ones that try it have documented race conditions and silent data loss. - [Built-in agents: what ships in the box](https://crabtalk.ai/blog/builtin-agents-landscape): We surveyed ten AI coding products to catalog their built-in agents. Three have real multi-agent delegation. The rest ship modes. - [Built-in tools: what your agent can reach](https://crabtalk.ai/blog/builtin-tools-landscape): We cataloged the built-in tools in ten AI coding products. Some ship 5, some ship 20+. The differences reveal what each product thinks an agent should do. - [Hermes memory: five layers, one learning loop](https://crabtalk.ai/blog/hermes-memory-system): We examined Hermes Agent's five-layer memory system — procedural skills, Honcho user modeling, FTS5 search — and asked what it costs to remember this much. - [Mem0: what three memory scopes actually cost](https://crabtalk.ai/blog/mem0-memory-architecture): We examined Mem0's extraction pipeline, conflict resolution, and benchmark claims. Smart memory management is real — but most agents don't need three scopes. - [Hermes Agent: what Nous Research built](https://crabtalk.ai/blog/hermes-agent-survey): We examined Hermes Agent's architecture — from Atropos RL training to persistent skill documents. Here's how it works and where it fits. - [When agents share a channel](https://crabtalk.ai/blog/agent-loop-prevention): We tested what happens when two agents join the same group chat. Most frameworks have no built-in answer for the infinite loop that follows. - [Context compaction in agent frameworks](https://crabtalk.ai/blog/context-compaction): We surveyed eight frameworks to understand who compacts what in multi-agent systems. The answer: each agent handles its own, and nobody coordinates. - [Tool permissions and the bash bypass problem](https://crabtalk.ai/blog/tool-call-permissions): If your agent has bash, do file-editing tools even matter for security? We surveyed eight frameworks and found the answer is architectural, not policy. - [How agents call agents](https://crabtalk.ai/blog/agent-calling-patterns): We surveyed eight frameworks to map who can call whom in agent teams — upward, lateral, recursive. Most are more restricted than expected. - [SOUL.md: brilliant idea, brittle implementation](https://crabtalk.ai/blog/soul-md-adoption): We analyzed OpenClaw's SOUL.md personality system — what it gets right about agent identity, and why static files break under production pressure. - [How AI frameworks control model thinking](https://crabtalk.ai/blog/thinking-control): We surveyed seven agent frameworks to understand who controls reasoning depth — the framework, the API, or nobody. The answers split three ways. - [What should an agent capability bench test?](https://crabtalk.ai/blog/agent-capability-benchmarks): A survey of existing agent benchmarks and 120+ questions we think a practical capability bench should answer. - [How developers configure their AI agents](https://crabtalk.ai/blog/agent-prompt-systems): A survey of CLAUDE.md adoption, MCP servers, skills, and hooks — and why your prompt system decays predictably from MVP to production. - [Sandboxing AI agents: beyond Docker and WASM](https://crabtalk.ai/blog/agent-sandbox-permissions): How Claude Code, Cursor, Codex, and Devin gate bash access — and why Docker, WASM, and OS sandboxes are all solving the wrong layer. - [Why multi-agent workflows fail in production](https://crabtalk.ai/blog/multi-agent-coordination): A survey of real coordination failures in Claude Code, Cursor, Devin, and the broader ecosystem — and what they reveal about agent architecture. - [How AI agents remember: a survey of persistent memory](https://crabtalk.ai/blog/persistent-agent-memory-research): We surveyed how Claude Code, ChatGPT, Cursor, Windsurf, and others implement persistent agent memory — storage formats, compaction, and retrieval. - [Plans vs tasks: how AI agents think before they act](https://crabtalk.ai/blog/plans-vs-tasks-agent-design): A survey of plan-then-execute patterns across Claude Code, Cursor, Devin, Windsurf, and Copilot — and what it means for autonomous agent design.