> ## Documentation Index
> Fetch the complete documentation index at: https://docs.agentium.in/llms.txt
> Use this file to discover all available pages before exploring further.
# Connect your stack
> Find model, speech, telephony, sandbox, telemetry, storage, toolkit, protocol, and device adapters.
export const IntegrationDirectory = () => {
const entries = [{
"name": "Anthropic",
"category": "Models",
"description": "Configure Claude, adaptive thinking, source content, and provider-owned continuation.",
"href": "/models/anthropic"
}, {
"name": "AWS Bedrock",
"category": "Models",
"description": "Use AWS Bedrock foundation models (Mistral, Amazon Nova, Llama, Cohere) with Agentium. Setup, factory, config, and multi-model support.",
"href": "/models/aws-bedrock"
}, {
"name": "AWS Claude",
"category": "Models",
"description": "Use Claude models through AWS Bedrock with Agentium. Native Anthropic SDK integration with AWS authentication.",
"href": "/models/aws-claude"
}, {
"name": "Azure AI Foundry",
"category": "Models",
"description": "Use open-source models (Phi, Llama, Mistral, Cohere) hosted on Azure AI Foundry with Agentium.",
"href": "/models/azure-foundry"
}, {
"name": "Azure OpenAI",
"category": "Models",
"description": "Configure OpenAI model deployments through Azure credentials and endpoints.",
"href": "/models/azure-openai"
}, {
"name": "Cohere",
"category": "Models",
"description": "Configure Cohere models, tool calling, and request options.",
"href": "/models/cohere"
}, {
"name": "Implement a model provider",
"category": "Models",
"description": "Normalize an upstream API into ModelProvider, register a factory, and verify generation and streaming contracts.",
"href": "/models/custom-provider"
}, {
"name": "DeepSeek",
"category": "Models",
"description": "Use DeepSeek reasoning and chat models with Agentium agents.",
"href": "/models/deepseek"
}, {
"name": "Google Gemini",
"category": "Models",
"description": "Use current Gemini text models and distinguish their configuration from Gemini Live.",
"href": "/models/google"
}, {
"name": "Jev (TypeSafe)",
"category": "Models",
"description": "Use the TypeSafe decision model for choices, boolean judgments, and scores.",
"href": "/models/jev"
}, {
"name": "Meta (Llama)",
"category": "Models",
"description": "Connect the Meta Llama model adapter to your application.",
"href": "/models/meta"
}, {
"name": "Mistral",
"category": "Models",
"description": "Configure Mistral models and their request options.",
"href": "/models/mistral"
}, {
"name": "Ollama (Local Models)",
"category": "Models",
"description": "Run local LLMs with Ollama and Agentium. No API key needed. Llama, CodeLlama, Mistral, and more.",
"href": "/models/ollama"
}, {
"name": "OpenAI-Compatible",
"category": "Models",
"description": "Connect to any OpenAI-compatible API endpoint — Together, Groq, Fireworks, OpenRouter, NVIDIA, LM Studio, and more.",
"href": "/models/openai-like"
}, {
"name": "OpenAI",
"category": "Models",
"description": "Configure the OpenAI text provider, GPT-6 tool calls, reasoning, and output contracts.",
"href": "/models/openai"
}, {
"name": "Perplexity",
"category": "Models",
"description": "Use Perplexity models with built-in web search, citations, and domain filtering in Agentium agents.",
"href": "/models/perplexity"
}, {
"name": "Vercel v0",
"category": "Models",
"description": "Configure the Vercel v0 model adapter for your application.",
"href": "/models/vercel"
}, {
"name": "Vertex AI",
"category": "Models",
"description": "Configure Gemini through Google Vertex AI and Google Cloud authentication.",
"href": "/models/vertex"
}, {
"name": "xAI (Grok)",
"category": "Models",
"description": "Configure the xAI Grok adapter and its model request options.",
"href": "/models/xai"
}, {
"name": "Calculator",
"category": "Toolkits",
"description": "Evaluate math expressions safely — no API key required.",
"href": "/toolkits/calculator"
}, {
"name": "Google Calendar",
"category": "Toolkits",
"description": "List, create, get, and delete Google Calendar events.",
"href": "/toolkits/calendar"
}, {
"name": "Code Interpreter",
"category": "Toolkits",
"description": "Execute JavaScript, Python, or TypeScript in a local subprocess with configured limits.",
"href": "/toolkits/code-interpreter"
}, {
"name": "Discord",
"category": "Toolkits",
"description": "Send messages, read messages, and list channels in Discord servers.",
"href": "/toolkits/discord"
}, {
"name": "DuckDuckGo",
"category": "Toolkits",
"description": "Search the web and get news using DuckDuckGo — no API key required.",
"href": "/toolkits/duckduckgo"
}, {
"name": "File System",
"category": "Toolkits",
"description": "Read, write, list, and inspect local files with path sandboxing.",
"href": "/toolkits/filesystem"
}, {
"name": "Git",
"category": "Toolkits",
"description": "Local git operations — status, diff, log, commit, branch.",
"href": "/toolkits/git"
}, {
"name": "GitHub",
"category": "Toolkits",
"description": "Interact with GitHub repos, issues, PRs, and file content.",
"href": "/toolkits/github"
}, {
"name": "Gmail",
"category": "Toolkits",
"description": "Send, search, and read emails via the Gmail API.",
"href": "/toolkits/gmail"
}, {
"name": "Google Sheets",
"category": "Toolkits",
"description": "Read, write, and append data in Google Sheets.",
"href": "/toolkits/google-sheets"
}, {
"name": "Google Workspace",
"category": "Toolkits",
"description": "Access 30+ Google Workspace APIs (Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin, and more) via the gws CLI.",
"href": "/toolkits/google-workspace"
}, {
"name": "Hacker News",
"category": "Toolkits",
"description": "Fetch top stories and user details from Hacker News — no API key required.",
"href": "/toolkits/hackernews"
}, {
"name": "HTTP / REST",
"category": "Toolkits",
"description": "Make HTTP requests through a toolkit with application-owned credentials and access policy.",
"href": "/toolkits/http"
}, {
"name": "Image Generation",
"category": "Toolkits",
"description": "Configure image generation and editing tools through the supported provider integration.",
"href": "/toolkits/image-generation"
}, {
"name": "Jev",
"category": "Toolkits",
"description": "Give a chat agent a judgment tool. Claude talks. Jev picks the label, yes/no, or score.",
"href": "/toolkits/jev"
}, {
"name": "Jira",
"category": "Toolkits",
"description": "Search, create, and update Jira issues from your agent.",
"href": "/toolkits/jira"
}, {
"name": "Notion",
"category": "Toolkits",
"description": "Search, read, and create pages in Notion databases.",
"href": "/toolkits/notion"
}, {
"name": "PageIndex",
"category": "Toolkits",
"description": "Vectorless, reasoning-based RAG — submit PDFs, build tree indexes, and retrieve via LLM reasoning.",
"href": "/toolkits/pageindex"
}, {
"name": "PDF",
"category": "Toolkits",
"description": "Extract text, metadata, and page content from PDF files.",
"href": "/toolkits/pdf"
}, {
"name": "Redis",
"category": "Toolkits",
"description": "Get, set, delete, list, and increment keys in Redis.",
"href": "/toolkits/redis"
}, {
"name": "S3 Cloud Storage",
"category": "Toolkits",
"description": "Upload, download, list, delete, and presign URLs for S3-compatible storage.",
"href": "/toolkits/s3"
}, {
"name": "Web Scraper",
"category": "Toolkits",
"description": "Extract page text and links through the web scraper toolkit.",
"href": "/toolkits/scraper"
}, {
"name": "Shell",
"category": "Toolkits",
"description": "Execute shell commands with timeout, output limits, and command allowlisting.",
"href": "/toolkits/shell"
}, {
"name": "Slack",
"category": "Toolkits",
"description": "Send and read messages, list channels, and reply in threads.",
"href": "/toolkits/slack"
}, {
"name": "SQL Database",
"category": "Toolkits",
"description": "Query databases (SQLite, PostgreSQL, MySQL) from your agent.",
"href": "/toolkits/sql"
}, {
"name": "Stripe",
"category": "Toolkits",
"description": "List charges, get customers, create refunds, list subscriptions, and get invoices.",
"href": "/toolkits/stripe"
}, {
"name": "Telegram",
"category": "Toolkits",
"description": "Send messages, photos, and read updates via the Telegram Bot API.",
"href": "/toolkits/telegram"
}, {
"name": "Web Search",
"category": "Toolkits",
"description": "AI-optimized web search using Tavily or SerpAPI backends.",
"href": "/toolkits/websearch"
}, {
"name": "WhatsApp",
"category": "Toolkits",
"description": "Send messages via Meta's WhatsApp Business Cloud API.",
"href": "/toolkits/whatsapp"
}, {
"name": "Wikipedia",
"category": "Toolkits",
"description": "Search and read Wikipedia articles — no API key required.",
"href": "/toolkits/wikipedia"
}, {
"name": "YouTube",
"category": "Toolkits",
"description": "Search YouTube videos and extract transcripts/captions.",
"href": "/toolkits/youtube"
}, {
"name": "DynamoDB Storage",
"category": "Storage",
"description": "AWS DynamoDB storage driver with on-demand billing",
"href": "/storage/dynamodb"
}, {
"name": "In-Memory Storage",
"category": "Storage",
"description": "InMemoryStorage driver—zero dependencies, no persistence. Default for agents when no storage is configured.",
"href": "/storage/in-memory"
}, {
"name": "MongoDB Storage",
"category": "Storage",
"description": "MongoDBStorage driver for document-based persistence. Requires mongodb. Call initialize() before use.",
"href": "/storage/mongodb"
}, {
"name": "MySQL Storage",
"category": "Storage",
"description": "MySQL-backed storage driver for sessions and memory",
"href": "/storage/mysql"
}, {
"name": "PostgreSQL storage",
"category": "Storage",
"description": "Persist namespaced JSON values with PostgresStorage, initialize its table, and close the host-owned pool.",
"href": "/storage/postgres"
}, {
"name": "Redis Storage",
"category": "Storage",
"description": "Redis-backed storage driver with optional TTL",
"href": "/storage/redis"
}, {
"name": "SQLite Storage",
"category": "Storage",
"description": "SqliteStorage driver for file-based persistence. Requires better-sqlite3. No initialize() needed.",
"href": "/storage/sqlite"
}, {
"name": "MCP Admin API",
"category": "Protocols",
"description": "Manage MCP server connections at runtime via REST endpoints — add, connect, disconnect, and remove MCP servers from a UI or API client.",
"href": "/mcp/admin"
}, {
"name": "MCP authorization helpers",
"category": "Protocols",
"description": "Validate issuer responses and select registration metadata in a host-owned OAuth flow.",
"href": "/mcp/auth-conformance"
}, {
"name": "MCP v2 and Tasks",
"category": "Protocols",
"description": "Use the split MCP client SDK with explicit protocol negotiation, credential audience, and task ownership.",
"href": "/mcp/v2"
}, {
"name": "A2A Client",
"category": "Protocols",
"description": "Connect to remote A2A agents — use them as tools, team members, or call them directly.",
"href": "/a2a/client"
}, {
"name": "API Gateway",
"category": "Protocols",
"description": "Merge local and remote endpoints under a single API surface",
"href": "/a2a/gateway"
}, {
"name": "Remote Teams",
"category": "Protocols",
"description": "Call remote team endpoints as local team objects",
"href": "/a2a/remote-team"
}, {
"name": "Remote Workflows",
"category": "Protocols",
"description": "Execute remote workflow endpoints",
"href": "/a2a/remote-workflow"
}, {
"name": "A2A Server",
"category": "Protocols",
"description": "Expose Agentium agents as A2A-compliant HTTP endpoints for cross-framework interoperability.",
"href": "/a2a/server"
}, {
"name": "A2A 1.0",
"category": "Protocols",
"description": "Expose authenticated A2A tasks with verified ownership and a bounded process-local lifecycle.",
"href": "/a2a/v1"
}, {
"name": "Bluetooth BLE",
"category": "Devices",
"description": "Scan, connect, and communicate with Bluetooth Low Energy devices.",
"href": "/edge/ble"
}, {
"name": "Camera",
"category": "Devices",
"description": "Capture photos and video from the Raspberry Pi camera module.",
"href": "/edge/camera"
}, {
"name": "Connect a device to your service",
"category": "Devices",
"description": "Implement the receiving HTTP contract, send bounded event batches, and review fetched configuration before applying it.",
"href": "/edge/edge-cloud"
}, {
"name": "GPIO",
"category": "Devices",
"description": "Control Raspberry Pi GPIO pins — read, write, watch edges, and software PWM.",
"href": "/edge/gpio"
}, {
"name": "Ollama on Edge",
"category": "Devices",
"description": "Model recommendations, auto-start, and pull helpers for running Ollama on Raspberry Pi.",
"href": "/edge/ollama-edge"
}, {
"name": "Monitor a device runtime",
"category": "Devices",
"description": "Observe resource pressure and watchdog timeouts, then implement admission and recovery in your host.",
"href": "/edge/runtime"
}, {
"name": "I2C Sensors",
"category": "Devices",
"description": "Read temperature, humidity, and pressure from I2C sensors like BME280.",
"href": "/edge/sensor"
}, {
"name": "Servo control",
"category": "Devices",
"description": "Configure ServoToolkit for hobby servo angles and sweeps through GPIO.",
"href": "/edge/servo"
}, {
"name": "System information",
"category": "Devices",
"description": "Read device health, process, and network information with SystemToolkit.",
"href": "/edge/system"
}, {
"name": "Voice sessions",
"category": "Runtimes",
"description": "Choose native realtime voice, a composable Agent pipeline, or buffered speech processing.",
"href": "/voice/overview"
}, {
"name": "Streaming voice with Agent",
"category": "Runtimes",
"description": "Connect recognizers, an Agent tool loop, synthesizers, and a host media transport.",
"href": "/voice/streaming"
}, {
"name": "Voice gateway",
"category": "Runtimes",
"description": "Connect browser audio to native voice sessions with bounded buffers and playback acknowledgements.",
"href": "/voice/gateway"
}, {
"name": "Outbound telephony",
"category": "Runtimes",
"description": "Manage authorized outbound call intents independently of your speech model and media transport.",
"href": "/voice/telephony"
}, {
"name": "Build a browser agent",
"category": "Runtimes",
"description": "Run a first browser task, inspect its actions, and choose DOM, vision, tool, and hosting integrations.",
"href": "/browser/overview"
}, {
"name": "Sandbox Agent",
"category": "Runtimes",
"description": "Manage an owned execution workspace with explicit local or remote lifecycle and cleanup.",
"href": "/features/sandbox-agent"
}, {
"name": "Cloud sandbox toolkits",
"category": "Runtimes",
"description": "Run code with owned E2B or Daytona sessions and explicit resource cleanup.",
"href": "/features/sandbox-toolkits"
}, {
"name": "Vector stores",
"category": "Retrieval",
"description": "Run a local retrieval example, then select a persistent vector backend with matching embeddings and explicit connection ownership.",
"href": "/knowledge/vector-stores"
}, {
"name": "Embeddings",
"category": "Retrieval",
"description": "Choose an embedding provider, keep model and dimensions consistent, and connect it to retrieval.",
"href": "/knowledge/embeddings"
}, {
"name": "Hybrid Search",
"category": "Retrieval",
"description": "Combine vector (semantic) and keyword (BM25) search with Reciprocal Rank Fusion for better retrieval accuracy.",
"href": "/knowledge/hybrid-search"
}, {
"name": "Reranking",
"category": "Retrieval",
"description": "Two-stage retrieval with Cohere, Voyage, Jina, or local ColBERT cross-encoders",
"href": "/features/reranking"
}, {
"name": "Twilio calls",
"category": "Telephony",
"description": "createTwilioCallProvider — outbound call control, provider credentials, status, and hangup.",
"href": "/telephony/twilio"
}, {
"name": "Telnyx calls",
"category": "Telephony",
"description": "createTelnyxCallProvider — outbound call control, provider credentials, status, and hangup.",
"href": "/telephony/telnyx"
}, {
"name": "SignalWire calls",
"category": "Telephony",
"description": "createSignalWireCallProvider — outbound call control, provider credentials, status, and hangup.",
"href": "/telephony/signalwire"
}, {
"name": "Vonage calls",
"category": "Telephony",
"description": "createVonageCallProvider — outbound call control, provider credentials, status, and hangup.",
"href": "/telephony/vonage"
}, {
"name": "Exotel calls",
"category": "Telephony",
"description": "createExotelCallProvider — outbound call control, provider credentials, status, and hangup.",
"href": "/telephony/exotel"
}, {
"name": "LiveKit SIP calls",
"category": "Telephony",
"description": "createLiveKitSipCallProvider — outbound call control, provider credentials, status, and hangup.",
"href": "/telephony/livekit"
}, {
"name": "OpenAI Realtime",
"category": "Speech and media",
"description": "OpenAIRealtimeProvider — Native speech conversation and tools.",
"href": "/voice/openai"
}, {
"name": "Google Gemini Live",
"category": "Speech and media",
"description": "GoogleLiveProvider — Native realtime audio and safe session resumption.",
"href": "/voice/google"
}, {
"name": "OpenAI transcription",
"category": "Speech and media",
"description": "OpenAIStreamingRecognizer — Streaming speech recognition and manual turn commit.",
"href": "/voice/openai"
}, {
"name": "ElevenLabs Scribe",
"category": "Speech and media",
"description": "ElevenLabsRecognizer — Streaming speech recognition.",
"href": "/voice/elevenlabs"
}, {
"name": "ElevenLabs text-to-speech",
"category": "Speech and media",
"description": "ElevenLabsSynthesizer — Stream-input synthesis with supported v2 models.",
"href": "/voice/elevenlabs"
}, {
"name": "ElevenLabs Speech Engine",
"category": "Speech and media",
"description": "createElevenLabsSpeechEngineHandler — Authenticated host engine callback for a VoiceBrain.",
"href": "/voice/elevenlabs"
}, {
"name": "Sarvam recognition",
"category": "Speech and media",
"description": "SarvamRecognizer — Preview Indian-language speech recognition.",
"href": "/voice/sarvam"
}, {
"name": "Sarvam text-to-speech",
"category": "Speech and media",
"description": "SarvamSynthesizer — Bulbul synthesis and explicit language support.",
"href": "/voice/sarvam"
}, {
"name": "LiveKit audio",
"category": "Speech and media",
"description": "LiveKitVoiceTransport — PCM media transport for a host-owned audio source.",
"href": "/voice/livekit"
}, {
"name": "Local sandbox",
"category": "Sandboxes",
"description": "SandboxAgent unix-local — trusted host code, owned workspace, bounded output.",
"href": "/sandbox/local"
}, {
"name": "E2B sandbox",
"category": "Sandboxes",
"description": "E2BSandbox and E2BSandboxToolkit — remote sessions, code tools, optional SDK, and cleanup.",
"href": "/sandbox/e2b"
}, {
"name": "Daytona sandbox",
"category": "Sandboxes",
"description": "DaytonaSandbox and DaytonaSandboxToolkit — remote sessions, code tools, optional SDK, and cleanup.",
"href": "/sandbox/daytona"
}, {
"name": "Console traces",
"category": "Telemetry",
"description": "ConsoleExporter — Console trace tree.",
"href": "/observability/exporters"
}, {
"name": "JSONL traces",
"category": "Telemetry",
"description": "JsonFileExporter — Append structured traces to a file.",
"href": "/observability/exporters"
}, {
"name": "Callback traces",
"category": "Telemetry",
"description": "CallbackExporter — Send sanitized traces to a host sink.",
"href": "/observability/exporters"
}, {
"name": "OpenTelemetry HTTP JSON",
"category": "Telemetry",
"description": "OTelExporter — Direct OTLP/HTTP JSON collector export.",
"href": "/observability/opentelemetry"
}, {
"name": "OpenTelemetry host bridge",
"category": "Telemetry",
"description": "OTelBridgeExporter — Borrow a host tracer and context binding.",
"href": "/observability/opentelemetry"
}, {
"name": "Langfuse OTLP",
"category": "Telemetry",
"description": "LangfuseOTLPExporter — Current Langfuse trace ingestion.",
"href": "/observability/langfuse"
}, {
"name": "Langfuse legacy",
"category": "Telemetry",
"description": "LangfuseExporter — Legacy compatibility and migration to OTLP.",
"href": "/observability/langfuse"
}];
const [query, setQuery] = useState("");
const [category, setCategory] = useState("All");
const categories = ["All", ...new Set(entries.map(entry => entry.category))];
const words = query.trim().toLowerCase().split(/[^\p{L}\p{N}]+/u).filter(Boolean);
const results = entries.filter(entry => {
const tokens = `${entry.name} ${entry.category} ${entry.description}`.toLowerCase().split(/[^\p{L}\p{N}]+/u);
return (category === "All" || entry.category === category) && words.every(word => tokens.some(token => token.startsWith(word)));
});
return
Search integrations
{categories.map(item => setCategory(item)}>{item} )}
{results.length} {results.length === 1 ? "integration" : "integrations"}{category !== "All" ? ` in ${category.toLowerCase()}` : ""}
{results.length === 0 && No matching integrations. {
setQuery("");
setCategory("All");
}}>Clear search and filters
}
;
};
Choose an integration by what your application needs to connect. Each guide covers its import path and configuration. Install the adapter's optional dependency explicitly and keep credentials in your host environment.