Memory for the agents you already have.
Every adapter wires the same five things into the framework's own extension points: the context before the answer, the turns recorded, the history as a tool, verification and handoff. None of them changes your model, your prompt or your vendor, and none fails the agent when Niadra is slow or down. Below, one page per platform, with the code of the published SDK.
Voice platforms
- LiveKit AgentsThe LiveKit agent starts the call knowing who is on the line.See how it wires in
- PipecatA frame processor between aggregators.user() and the LLM.See how it wires in
- VapiVapi posts every server message to one URL, and one handler answers the ones that matter: the context on the assistant-request, while the assistant is not answering yet; the history tools on tool-calls; the transfer destination; and the end-of-call-report as turns.See how it wires in
- Retell AIRetell reaches your server three ways, and the adapter answers each: the inbound call webhook with the context as a dynamic variable, the Retell LLM's custom functions with the history tools and the agent webhook with the call's events.See how it wires in
- ElevenLabs Agents PlatformA call to an ElevenLabs agent reaches your server through three webhooks, and the adapter answers each: the initiation one hands over the context as a dynamic variable, the server tools run the history for whoever is on the line, and the post-call one records the transcript and ends the conversation.See how it wires in
Channels
- TwilioThe adapter reads the Programmable Voice and Messaging webhooks, SMS and WhatsApp, and in TypeScript the Conversations ones: it checks X-Twilio-Signature, finds the customer, the call or thread id and what the carrier attested, and records the inbound turn.See how it wires in
- WhatsApp Cloud APIThe WhatsApp Cloud API adapter only translates: it checks the signature of Meta's webhook, extracts each message with the customer's identifier and the item ready to record, and records the agent's reply from the send API's answer.See how it wires in
Agent frameworks
- OpenAI Agents SDKIn Python, NiadraAgentsMemory gives you run_config(), the filter that injects the context before every model call, hooks, which record the turns, and tools.See how it wires in
- LangGraphNiadraMiddleware wraps every model call of the agent built by langchain.agents.create_agent: the pack goes into the system message, the turns are recorded and the history tools come along.See how it wires in
- LangChainIn Python, with langchain-core only: context_runnable() places the context in the prompt's messages, NiadraCallbackHandler records the turns and history_tools() hands over the kit as StructuredTools.See how it wires in
- CrewAINiadraCrew gives you before_kickoff, which puts the pack in the crew's inputs as niadra_context, after_kickoff, which records the final answer, and tools.See how it wires in
- Vercel AI SDKniadraMiddleware(session) is a language model middleware for wrapLanguageModel({ model, middleware }): the AI SDK's idiomatic path, which works with every provider without a separate provider package.See how it wires in
Automation
Agno, Mastra, Google ADK, Strands, Pydantic AI, LlamaIndex, Semantic Kernel, Bedrock, Anthropic, Google GenAI, LiteLLM, Dify, Flowise, Langflow and the others are in the documentation, with the same design.
All 36 integrations, in the documentation(opens docs.niadra.com)Tell us what you are building.
A work email and two lines about your agents are enough. The people who write the code reply, with an early-access proposal for your case.
Rather tell us more about your company? Use the full form