# kunchenguid/axi

Design principles for agent ergonomics. Higher accuracy with lower token cost than both MCP and regular CLI.

Repository: https://github.com/kunchenguid/axi
Canonical: https://ross.abutalabs.com/products/axi
Homepage: https://axi.md/
Language: TypeScript
License: MIT
License Family: permissive
Last push: 2026-08-26T19:31:37+00:00

## Health v2 (maintenance only)
Score: 81/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 98, longevity 11
- inputs: {"age_days": 165, "days_push": 7, "days_rel": 13, "gap_med": 10.0, "n_releases_24m": 13}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1921, forks 146 (observed 2026-08-28T04:05:53.955494+00:00)

## What it is
AXI (Agent eXperience Interface) is a set of 10 design principles plus reference CLI tools for building agent-ergonomic applications that treat token budget as a first-class constraint. It ships reference implementations like gh-axi and chrome-devtools-axi that outperform MCP and human-oriented CLIs on cost, speed, and task success in published benchmarks.

## Use cases
- reduce token cost when agents call CLI tools
- replace MCP servers with cheaper agent-native interfaces
- build agent-ergonomic CLI tools following design principles
- benchmark agent tool interfaces for success rate and cost
- automate GitHub operations from an AI agent
- automate browser tasks with lower overhead than MCP

## When to choose
- you are building CLI tools primarily consumed by LLM agents
- you want lower token cost and fewer turns than MCP tool protocols
- you need validated, benchmarked agent tooling patterns
- you want ready-made agent-friendly wrappers for GitHub or Chrome DevTools

## When to avoid
- you need a mature, widely-adopted standard with broad ecosystem support
- your tools are primarily for human users rather than agents
- you require MCP-specific features like transport-level integrations
- you need guaranteed results across models other than those benchmarked

## Facets
- artifact type: framework
- maturity: active
- function: cli, agent-framework, developer-tools
- domain: developer-tools, large-language-models
- platform: cli, cross-platform
- tags: agent-ergonomics, token-efficiency, mcp-alternative, agent-native-cli, design-principles, benchmarking, ai-agents, command-line, nodejs

## Member repositories
- kunchenguid/axi (main) score 81

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:53.955494+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 2026-08-30T03:09:42.358369+00:00, confidence not recorded.
  - readme: https://github.com/kunchenguid/axi (fetched 2026-08-28T04:05:53.955494+00:00, sha f6bec72f3b4c)
  - homepage: https://axi.md/ (fetched 2026-08-29T10:49:22.739170+00:00, sha 82919f7bb2ab)
- Data as of 2026-08-30T08:39:29.467469+00:00.