vcoderun/autobench

최근 기록된 소스 활동: 2026년 10월 14일 18:35

SkillsMP 카탈로그 업데이트: 2026년 11월 9일 14:22

수집된 skills: 1

GitHub 스타: 0
GitHub 포크: 0

AI 어시스턴트로 설치

이 프롬프트를 복사해 사용 중인 AI 어시스턴트에 입력하세요.

Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/vcoderun/autobench.

저장소 README

README 원본 파일: README.md

README 원본 보기

README 전체 펼치기

# Autobench When an application changes, you need to know whether the new version is more correct, faster, cheaper, or more reliable. Teams usually answer that question with a custom benchmark script: a small program that runs representative inputs, checks the outputs, collects measurements, and prints a comparison. As the application grows, these scripts are repeatedly rewritten and their results become difficult to reproduce or compare. We built Autobench to solve that problem. Define the cases, variants, task, scores, and reports once in YAML or Python. Autobench runs the experiment, collects the evidence, records exactly what happened, and lets you replay, report, compare, or export the result without rebuilding that infrastructure for every application. It is a YAML-first Python framework for AI and non-AI systems: - deterministic dataset x variant execution - sync and async application tasks - semantic observations, checks, measurements, artifacts, and ABP traces - built-in and custom scoring, cost derivation, policies, and paired baselines - native Pydantic AI, OpenAI, OpenAI Agents, and HTTPX instrumentation - explicit and automatic prompt/tool/schema/agent asset lineage - immutable YAML records, replay, Rich reports, comparisons, and exports - optional immutable-record export to OTLP-compatible telemetry backends ## Install ```bash uv add autobench ``` For native SDK instrumentation: ```bash uv add 'autobench[instrumentation]' ``` For outbound OTLP HTTP/protobuf export: ```bash uv add 'autobench[otlp]' ``` ## First Run ```bash autobench validate examples/minimal/autobench.yaml autobench run examples/minimal/autobench.yaml --record /tmp/autobench-minimal autobench replay /tmp/autobench-minimal autobench report /tmp/autobench-minimal ``` A task is a normal sync or async callable: ```python from autobench import Case, RunContext def run(ctx: RunContext, case: Case) -> Result: mode = ctx.factor("mode") with ctx.span("subject", kind="workflow") as span: result = application(case.input, mode=mode) span.set_output(result) return result ``` The YAML spec owns reusable benchmark infrastructure: cases, variants, scoring, derivation, policies, instrumentation, and reports. ## Examples | Directory | Demonstrates | | --- | --- | | `examples/minimal` | Inline cases, variants, exact scoring, report and comparison | | `examples/basic` | File dataset, spans, checks, artifacts and failure visibility | | `examples/mid` | Semantic usage, pricing, cost, policies and distributions | | `examples/advanced` | Repeated measurement and paired-baseline speedup | | `examples/pydantic_ai` | Live layered instrumentation and automatic asset discovery | | `examples/automatic_assets` | Offline Pydantic AI and custom SDK behavioral lineage | | `examples/abp_*` | Manual, concurrent, streaming, Agents and replay protocol flows | | `examples/otlp_export` | Offline immutable ABP record to OTLP mapping | | `examples/codemode` | Migration of a real external benchmark runner | Run the offline release matrix: ```bash make examples ``` ## Documentation Full documentation: [vcoderun.github.io/autobench](https://vcoderun.github.io/autobench/) - [Installation](https://vcoderun.github.io/autobench/installation/) - [First Benchmark](https://vcoderun.github.io/autobench/getting-started/) - [Use Cases](https://vcoderun.github.io/autobench/use-cases/) - [Architecture](https://vcoderun.github.io/autobench/architecture/) - [YAML Spec](https://vcoderun.github.io/autobench/yaml-spec/) - [Python API](https://vcoderun.github.io/autobench/python-api/) - [Autobench Protocol](https://vcoderun.github.io/autobench/instrumentation-and-traces/) - [OTLP Export](https://vcoderun.github.io/autobench/otlp-export/) LLM-readable indexes are available at [`llms.txt`](https://vcoderun.github.io/autobench/llms.txt) and [`llms-full.txt`](https://vcoderun.github.io/autobench/llms-full.txt). ## Development ```bash uv sync --extra dev --extra instrumentation --extra openai-agents --extra otlp make prod make pre-commit ``` The release gate enforces Python 3.11-3.14, strict lint and typing, strict documentation builds, offline examples, and 100% source line and branch coverage.

저장소 README를 불러오는 중…

이 저장소의 skills

분류 대기 중

skill

직업 분류

설명

업데이트

autobench

미분류

Design, implement, run, debug, review, or document Autobench benchmarks and integrations. Use for Autobench YAML specs, Python Benchmark builders, task(ctx, case) functions, datasets and variants, semantic metrics, scoring and derivation, policies, ABP…

원문 언어: 영어

2026-10-14

수집된 skill 1개 중 1개를 표시합니다.