# katanemo/plano

Plano is an AI-native proxy server and data plane for agentic apps. Smart LLM routing, observability, agent orchestration, and guardrails so you stay focused on your agents core logic.

Repository: https://github.com/katanemo/plano
Canonical: https://ross.abutalabs.com/products/plano
Homepage: https://planoai.dev
Language: Rust
License: Apache-2.0
License Family: permissive
Topics: gateway, generative-ai, llm-inference, llms, prompt, proxy, proxy-server, llmops, openai, routing, ai-gateway, llm-gateway, llm-routing, envoy, envoyproxy, ai-gateway-support, llm-proxy
Last push: 2026-08-19T19:29:08+00:00

## Health v2 (maintenance only)
Score: 85/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 98, release rhythm 86, longevity 56
- inputs: {"age_days": 785, "days_push": 14, "days_rel": 14, "gap_med": 7.0, "n_releases_24m": 73}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7017, forks 488 (observed 2026-08-28T04:09:53.384971+00:00)

## What it is
Plano is an AI-native proxy server and data plane for agentic applications, built on Envoy by its core contributors. It centralizes agent routing and orchestration, LLM model routing, observability signals and traces, and guardrail filter chains so developers can ship agents to production without bespoke middleware.

## Use cases
- route requests to the right agent in a multi-agent system
- proxy and load-balance across multiple LLM providers with model aliases
- add guardrails and moderation filters to LLM traffic
- capture traces and agentic signals for observability
- standardize LLM access behind an OpenAI-compatible gateway
- turn prompts into deterministic API function calls
- deploy agents on-premises with centralized policy control

## When to choose
- you run multiple agents or LLM providers and need unified routing and failover
- you want zero-code observability, tracing, and metrics across agentic apps
- you need centralized guardrails, moderation, and access policies
- you want framework-agnostic infrastructure that works with any language or AI framework
- you need on-premises or self-hosted delivery infrastructure for regulated environments

## When to avoid
- you have a single-model, single-agent prototype with no production concerns
- you need a client-side SDK embedded in application code rather than a network proxy
- you require Windows-native deployment without Docker
- you want a full agent framework with built-in agent code abstractions rather than a data plane

## Facets
- artifact type: service
- maturity: active
- function: proxy, api-gateway, routing, middleware, tracing, monitoring, llm-inference, agent-framework, security
- domain: large-language-models, developer-tools, infrastructure-as-code, self-hosted, apis, networking
- platform: self-hosted, cloud, cli
- tags: llm-gateway, ai-gateway, llm-routing, envoy, agentic-apps, guardrails, llmops, dataplane, openai-compatible, rust, ai-agents, linux, macos, docker

## Member repositories
- katanemo/plano (main) score 85

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:53.384971+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:40:42.076303+00:00, confidence not recorded.
  - readme: https://github.com/katanemo/plano (fetched 2026-08-28T04:09:53.384971+00:00, sha 97a9c7abd4d0)
  - homepage: https://planoai.dev (fetched 2026-08-29T08:36:34.445413+00:00, sha 6bddb167e03d)
  - site_page: https://docs.planoai.dev/get_started/quickstart (fetched 2026-08-29T08:36:34.457655+00:00, sha 457b107a08d0)
  - site_page: https://docs.planoai.dev (fetched 2026-08-29T08:36:34.459930+00:00, sha 9596507c55f1)
- Data as of 2026-08-30T08:39:29.467469+00:00.