By ajshedivy
Interact with AgentOS API endpoints — manage agents, teams, sessions, traces, evals, knowledge, memory, and workflows remotely
Interact with AgentOS Agent API endpoints. For standard operations (listing agents, running agents, streaming), use the provided CLI script first. Only write custom Python when the script cannot handle the use case (e.g., structured output, dependencies, cancellation, chaining multiple calls). Trigger when: running agents remotely, listing agents, creating agent tests, or asking things like "kick off a run in my test agent" or "what agents do I have configured?"
Interact with AgentOS Evals API endpoints. For standard operations (listing evals, running accuracy/performance evals, getting eval details), use the provided CLI script first. Only write custom Python when the script cannot handle the use case (e.g., advanced filtering, update/delete operations, chaining evals, agent-as-judge). Trigger when: running evaluations, listing eval runs, benchmarking agents, or asking things like "run an accuracy eval on my agent" or "show me the latest eval results."
Interact with AgentOS Knowledge API endpoints. For standard operations (listing content, uploading files, searching, deleting), use the provided CLI script first. Only write custom Python when the script cannot handle the use case (e.g., pagination, updating metadata, bulk workflows, get config). Trigger when: uploading documents, searching the knowledge base, listing content, or asking things like "upload this md file to my knowledge base" or "search my docs for X."
Interact with AgentOS Memory API endpoints. For standard operations (listing, creating, updating, deleting memories, searching, topics), use the provided CLI script first. Only write custom Python when the script cannot handle the use case (e.g., advanced filtering, stats, chaining multiple calls, integration tests). Trigger when: managing user memories, writing scripts to manage user preferences, creating memory tests, or asking things like "what memories does this user have?" or "add a preference for dark mode."
Interact with AgentOS Session API endpoints. For standard operations (listing, creating, renaming, deleting sessions, viewing runs), use the provided CLI script first. Only write custom Python when the script cannot handle the use case (e.g., persistent conversations, clearing history, bulk deletion, advanced filtering). Trigger when: managing sessions remotely, inspecting session history, creating session tests, or asking things like "find me all sessions from the researcher agent" or "get the latest runs from session X."
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Cookbook examples for Agno with configurable model selection. Switch models via .env instead of editing code.
# Clone and setup
git clone <repo-url> && cd agno-cookbook
cp .env.example .env
# Install dependencies + your provider SDK
uv sync --extra quickstart --extra anthropic
# Run an example
uv run python cookbook/00_quickstart/agent_with_tools.py
Edit .env with your API key and preferred model:
AGNO_MODEL=anthropic:claude-sonnet-4-20250514
ANTHROPIC_API_KEY=sk-ant-...
Set AGNO_MODEL in .env using the format provider:model_id. Agno supports 40+ providers.
You must install the provider's SDK via uv sync --extra <provider>:
AGNO_MODEL | Install command | API key env var |
|---|---|---|
openai:gpt-4o | uv sync --extra openai | OPENAI_API_KEY |
anthropic:claude-sonnet-4-20250514 | uv sync --extra anthropic | ANTHROPIC_API_KEY |
google:gemini-3-flash-preview | uv sync --extra google | GOOGLE_API_KEY |
groq:llama-3.3-70b-versatile | uv sync --extra groq | GROQ_API_KEY |
mistral:mistral-large-latest | uv sync --extra mistral | MISTRAL_API_KEY |
cohere:command-r-plus | uv sync --extra cohere | COHERE_API_KEY |
ollama:llama3.2 | uv sync --extra ollama | (none — runs locally) |
You can combine multiple extras:
uv sync --extra quickstart --extra anthropic # Quickstart tools + Anthropic
uv sync --extra all-models # All major provider SDKs
The agentos-serve package lets you spin up AgentOS from any Python file containing Agno agents, teams, or workflows.
uv pip install -e agentos-serve
Or install it via the serve extra (non-editable, from PyPI):
uv sync --extra serve
# Serve a single file
agentos-serve cookbook/02_agents/01_quickstart/basic_agent.py
# Serve all agents in a directory
agentos-serve cookbook/02_agents/01_quickstart/
# The cookbook also provides an `agentos` alias
uv run agentos cookbook/02_agents/01_quickstart/
See agentos-serve/README.md for full CLI options and extras.
cookbook/
00_quickstart/ # Start here
01_demo/ # Full demo app
02_agents/ # Agent patterns
03_teams/ # Multi-agent teams
04_workflows/ # Workflow orchestration
05_agent_os/ # Agent OS examples
06_storage/ # Persistence
07_knowledge/ # Knowledge bases & embedders
08_learning/ # Self-learning agents
09_evals/ # Evaluation patterns
10_reasoning/ # Reasoning strategies
11_memory/ # Memory management
90_models/ # Provider-specific demos (hardcoded models)
91_tools/ # Tool integrations
92_integrations/ # Third-party integrations
93_components/ # UI components
Note:
cookbook/90_models/contains provider-specific examples with hardcoded model classes. These are intentionally not configurable — they demonstrate provider-specific features.
The plugins/ directory contains agent skill packs that provide domain-specific guidance, CLI scripts, and API reference patterns for building with the Agno ecosystem.
Install skills as Claude Code plugins:
# Add the marketplace from GitHub
claude plugin marketplace add ajshedivy/agno-cookbook
# Install the skill packs
claude plugin install agno-framework
claude plugin install agno-agentos-api
Once connected, skills activate automatically based on context — asking Claude to "create an Agno agent with tools" triggers agno-agent, and "list my agents on AgentOS" triggers agentos-api-agents.
agno-framework — Agno SDK Development13 skills covering all major Agno framework areas:
| Skill | Description |
|---|---|
agno-agent | Agent creation, configuration, and tool binding |
agno-team | Multi-agent team orchestration |
agno-workflow | Workflow orchestration patterns |
agno-agentos | AgentOS server setup and deployment |
agno-tools | Tool development and integration |
agno-knowledge | Knowledge bases and RAG |
agno-memory | Memory and session management |
agno-storage | Persistence backends |
agno-models | Model providers and configuration |
agno-reasoning | Reasoning strategies |
agno-multimodal | Image, audio, and video support |
agno-guardrails | Input/output validation |
agno-test | Testing and evaluation patterns |
agno-agentos-api — AgentOS Client API8 skills for interacting with a running AgentOS instance via the AgentOSClient SDK. Each skill includes a CLI script for common operations and reference docs for advanced usage.
npx claudepluginhub ajshedivy/agno-cookbook --plugin agno-agentos-apiBuild AI agents, teams, and workflows with the Agno framework — includes tools, knowledge, memory, storage, guardrails, reasoning, and multimodal support
24 agent skills with 215+ tools for IBM i SQL Services — full coverage of work management, database, security, storage, communication, and system administration via QSYS2 and SYSTOOLS services. Driven by the @ibm/ibmi-cli.
Operate a running AgentOS server from the command line with agno-cli — agents, teams, workflows, traces, sessions, memories, knowledge, evals, approvals, schedules, and config.
Official Agno AI agent framework skill - build production-ready agents, multi-agent teams, workflows, MCP integrations, and deploy with AgentOS.
Hephaestus meta-agent team for creating Agentlas agents and teams, packaging existing agents, and opening the local ontology GUI.
Connect Claude Code to the Viche agent network — discover, message, and collaborate with other AI agents in real-time
A plugin for Claude Code agent team workflows — session viewing, analysis, and best practice evaluation
Agent team management - create, expand, and cleanup teams for worktree sessions
Production ADK orchestrator for A2A protocol and multi-agent coordination on Vertex AI