# crabtalk.ai

> AI-optimized mirror of crabtalk.ai containing 50 pages totalling 13,067 words of clean markdown content, structured data, and semantic HTML. Original source: https://crabtalk.ai/. Last updated: 2026-05-01T01:34:35.258Z.

Each page is available as HTML (with JSON-LD structured data) and Markdown (text-only, ideal for LLMs and RAG).

## Homepage

- [Where agentsLAND.](/site-root.html): Stop checking ten places. Build one agent. It reads what mattered, posts what it shipped, answers when asked. Open-source substrate, platform in flight. (87 words)

## Articles & Blog Posts

- [blog/async-compaction-challenges/index.html](/blog/async-compaction-challenges/index.html) (1 words)
- [blog/agent-sandbox-permissions/index.html](/blog/agent-sandbox-permissions/index.html) (1 words)
- [blog/agent-loop-prevention/index.html](/blog/agent-loop-prevention/index.html) (1 words)
- [blog/agent-capability-benchmarks/index.html](/blog/agent-capability-benchmarks/index.html) (1 words)
- [blog/agent-prompt-systems/index.html](/blog/agent-prompt-systems/index.html) (1 words)
- [blog/agent-calling-patterns/index.html](/blog/agent-calling-patterns/index.html) (1 words)
- [Management](/docs/crabllm/features/management/index.html): Dynamic provider and key management via crabctl CLI, admin API, and OpenAPI docs. (548 words)
- [Hub](/docs/crabtalk/hub/index.html): Discover and install MCP servers, agents, and commands from the CrabTalk hub. (334 words)
- [Events](/docs/crabtalk/events/index.html): Real-time event stream — monitor every agent, every tool call, every session from one connection. (280 words)
- [Docker](/docs/crabllm/docker/index.html): Run CrabLLM in Docker with persistent storage and manage it with crabctl from the host. (347 words)
- [Providers](/docs/crabtalk/providers/index.html): How CrabTalk routes model requests — API standards, provider config, and supported providers. (444 words)
- [Protocol](/docs/crabtalk/protocol/index.html): The wire protocol between CrabTalk clients and the daemon — protobuf over UDS or TCP. (666 words)
- [Configuration](/docs/crabllm/configuration/index.html): Full reference for crabllm.toml — providers, virtual keys, aliases, pricing, extensions, and storage. (379 words)
- [Extensions](/docs/crabllm/features/extensions/index.html): Request pipeline hooks — caching, rate limiting, usage tracking, budget enforcement, and logging. (345 words)
- [Sessions](/docs/crabtalk/sessions/index.html): Persistent conversations — append-only JSONL, auto-compaction, crash-safe. (392 words)
- [Providers](/docs/crabllm/providers/overview/index.html): Supported LLM providers, translation types, and endpoint support matrix. (245 words)
- [Just the DAEMON.](/oss/crabtalk/index.html): A 5 MB agent daemon that hides nothing. Stream every event, manage sessions, hot-swap commands — all from your client. (584 words)
- [Architecture](/docs/crabllm/architecture/index.html): Design principles, workspace layout, and request flow through the CrabLLM gateway. (470 words)
- [Authentication](/docs/crabllm/features/auth/index.html): Virtual API keys with per-key model access control, usage tracking, and no-auth mode. (240 words)
- [CrabDash Gateway API](/docs/crabdash/gateway-api/index.html): Use CrabDash's local OpenAI-compatible and Anthropic-compatible endpoints from any app, SDK, or CLI tool. (435 words)
- [Getting Started](/docs/crabllm/getting-started/index.html): Install CrabLLM, configure providers, and send your first request in under a minute. (249 words)
- [Agents](/docs/crabtalk/agents/index.html): How CrabTalk agents work — configuration, system prompts, delegation, and scoping. (439 words)
- [Introduction](/docs/crabllm/introduction/index.html): High-performance LLM API gateway in Rust — one API format, many providers, sub-millisecond overhead. (265 words)
- [MCP Servers](/docs/crabtalk/mcp-servers/index.html): Connect external tool servers to CrabTalk — stdio or HTTP, auto-discovered, hot-reloadable. (393 words)
- [Routing](/docs/crabllm/features/routing/index.html): Weighted random selection, exponential backoff retry, automatic provider fallback, and timeouts. (390 words)
- [OpenAI](/docs/crabllm/providers/openai/index.html): Configure OpenAI and any OpenAI-compatible API — Groq, Together AI, vLLM, and custom endpoints. (187 words)
- [Streaming](/docs/crabllm/features/streaming/index.html): SSE streaming across all providers — proxied without buffering, with per-chunk extension hooks. (246 words)
- [Ollama](/docs/crabllm/providers/ollama/index.html): Configure Ollama for local model inference — no API key, OpenAI-compatible pass-through. (146 words)
- [AWS Bedrock](/docs/crabllm/providers/bedrock/index.html): Configure AWS Bedrock — Converse API translation with SigV4 request signing. (165 words)
- [Skills](/docs/crabtalk/skills/index.html): Prompt-level behavior for agents — Markdown files with YAML frontmatter, discoverable and composable. (305 words)
- [Commands](/docs/crabtalk/commands/index.html): CrabTalk commands — cargo-style subcommand dispatch for telegram, search, and custom tooling. (471 words)
- [Tutorials](/docs/crabtalk/tutorials/index.html): Step-by-step guides for setting up and extending CrabTalk. (166 words)
- [One API.  Every model.](/oss/crabllm/index.html): Production LLM proxy in Rust. Sub-millisecond overhead, weighted routing, automatic failover, response caching, budget enforcement. Single binary. (355 words)
- [REPL](/docs/crabtalk/repl/index.html): Interactive chat with your agent — slash commands, tab completion, streaming output. (306 words)
- [Storage](/docs/crabllm/features/storage/index.html): Pluggable storage backends — memory, SQLite, and Redis for extension state persistence. (185 words)
- [Configuration](/docs/crabtalk/config/index.html): Configure CrabTalk — config.toml, models, MCP servers, and agents. (255 words)
- [Anthropic](/docs/crabllm/providers/anthropic/index.html): Configure Anthropic Claude — full request/response translation including system messages, tool calling, and streaming. (183 words)
- [Google Gemini](/docs/crabllm/providers/google/index.html): Configure Google Gemini — full translation including system instructions, function calling, and streaming. (148 words)
- [Azure OpenAI](/docs/crabllm/providers/azure/index.html): Configure Azure OpenAI — deployment-based URL rewriting and API version handling. (150 words)
- [Cron](/docs/crabtalk/cron/index.html): Daemon-level cron — run skills on a schedule, even when no client is connected. (307 words)
- [Memory](/docs/crabtalk/memory/index.html): Local-only agent memory — remember, recall, and forget, all on your machine. (271 words)
- [Libraries](/docs/crabllm/libraries/index.html): Standalone Rust crates for local inference — MLX (Apple Silicon) and llama.cpp (cross-platform). (149 words)
- [CrabDash Free Provider Keys](/docs/crabdash/provider-keys/index.html): Get free LLM API keys to use with CrabDash — no credit card, no waitlist. (373 words)
- [Getting Started with CrabDash](/docs/crabdash/getting-started/index.html): Download CrabDash, add your first LLM provider key, and route requests through your personal local gateway. (285 words)
- [Quickstart](/docs/crabtalk/quickstart/index.html): Run your first CrabTalk agent in under a minute — start the daemon and chat. (203 words)
- [Installation](/docs/crabtalk/installation/index.html): Install CrabTalk on macOS, Linux, or Windows with one command, or build from source with Cargo. (124 words)
- [CrabDash Local Models](/docs/crabdash/local-models/index.html): Run ~700 MLX community models locally on Apple Silicon — no Docker, no Ollama, no separate process. (266 words)
- [Auth](/docs/crabtalk/auth/index.html): Manage providers and MCP servers with the interactive auth TUI. (166 words)
- [Every model.  One menubar.](/apps/crabdash/index.html): Every request you send, every model you run, every dollar you spend — visible in your menubar. A macOS-first LLM gateway dashboard. (117 words)

## Resources

- [Full Page Index](/index.html): Browse all cached pages with rich metadata
- [About This Cache](/content/about.html): Methodology, technical details, and usage guidelines
- [XML Sitemap](/content/sitemap.xml): Machine-readable sitemap for crawler discovery
- [Robots.txt](/content/robots.txt): Crawler directives