# microsoft/waza

CLI / Framework for Agent Skills - create, test, measure and improve skill quality and effectiveness

Repository: https://github.com/microsoft/waza
Canonical: https://ross.abutalabs.com/products/microsoft-waza
Homepage: https://microsoft.github.io/waza/
Language: Go
License: MIT
License Family: permissive
Last push: 2026-08-24T16:55:03+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 13
- inputs: {"age_days": 187, "days_push": 9, "days_rel": 14, "gap_med": 0.0, "n_releases_24m": 45}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1280, forks 79 (observed 2026-08-28T04:04:13.688641+00:00)

## What it is
Waza is a Go CLI from Microsoft for evaluating AI agent skills through structured benchmarks. Users define test cases and validation rules in YAML, run them against different AI models, and compare results via graders, metrics, and a web dashboard.

## Use cases
- evaluate ai agent skills with benchmarks
- compare llm performance across models
- run regression tests on agent prompts in ci
- scaffold an eval suite for a new skill
- validate ai task completion quality with graders
- track skill effectiveness over time with a dashboard

## When to choose
- you build or maintain AI agent skills and need measurable quality metrics
- you want YAML-defined evals runnable locally and in GitHub Actions
- you need cross-model comparison of agent behavior

## When to avoid
- you need general-purpose LLM app tracing or observability rather than skill benchmarking
- you want a GUI-first evaluation platform with no CLI workflow
- you need training or fine-tuning tooling rather than evaluation

## Facets
- artifact type: cli-tool
- maturity: active
- function: benchmarking, testing, cli, agent-framework, developer-tools, ci-cd, data-visualization
- domain: artificial-intelligence, developer-tools, testing, large-language-models
- platform: windows, cli, go
- tags: ai-agent-skills, eval-suite, yaml-benchmarks, model-comparison, graders, github-copilot, ci-cd-integration, microsoft, ai-agents, linux, macos

## Member repositories
- microsoft/waza (main) score 81

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:13.688641+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 2026-08-30T05:02:37.934854+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/waza (fetched 2026-08-28T04:04:13.688641+00:00, sha d0ca31520cb5)
  - homepage: https://microsoft.github.io/waza/ (fetched 2026-08-29T12:13:09.424077+00:00, sha 59132ac56236)
- Data as of 2026-08-30T08:39:29.467469+00:00.