Claude 4.5 Sonnet

Abe Diaz / abe238 公开的代码代理与软件工程案例，来源为 公开代码库，复核于 2026-06-27T10:13:27Z。 Model Atlas 将它标记为 A 类证据，因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。

任务：这是一个围绕软件工程任务执行的真实任务，公开材料可以回溯到具体使用者和具体产物。 原始资料写作：Build an open-source Applied AI Product Manager resume analyzer for job seekers and hiring teams, using a six-pillar AI PM evaluation framework to score PDF, DOC, and DOCX resumes and produce structured hiring-readiness…

公开产物：公开材料提供公开代码、README 或项目配置，可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作：The public project publishes a runnable analyzer and product site; its README states it generates per-candidate Markdown, HTML, and JSON reports with 0-10 scoring by pillar, evidence, level assessment, and model-selecta…

模型作用：Claude 4.5 Sonnet 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作：Claude Sonnet 4.5 is explicitly listed as an Anthropic option and default Anthropic model for the analyzer; the CLI model table binds claude-sonnet-4-5-20250929 to complex resume-analysis work and the README documents r…

A 类理由：有具体使用者、具体任务、公开原始证据和可访问产物，可公开核验。

风险边界：当前判断基于公开材料；若产物下线、仓库变更或模型参与比例仅来自作者自述，需要在引用前重新复核。 原始资料写作：The tool is multi-provider and defaults to OpenAI unless configured; accepted because the README and CLI code explicitly bind the Anthropic pathway to claude-sonnet-4-5-20250929 and provide a public artifact/product for…

原始记录：Abe Diaz built an AI PM resume analyzer that can use Claude Sonnet 4.5 for six-pillar hiring evaluations

abe238 / AI PM Resume Analyzer 公开的代码代理与软件工程案例，来源为 公开代码库，复核于 2026-06-27T10:12:11Z。 Model Atlas 将它标记为 A 类证据，因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。

任务：这是一个围绕软件工程任务执行的真实任务，公开材料可以回溯到具体使用者和具体产物。 原始资料写作：Analyze Applied AI PM resumes from PDF/DOC/DOCX files against a 2026 six-pillar framework covering technical skills, product thinking, AI/ML knowledge, communication, strategic thinking, and execution; users can run the…

公开产物：公开材料提供公开代码、README 或项目配置，可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作：The tool generates structured resume evaluations with 0-10 pillar scores, evidence quotes, total score out of 60, decision labels such as Strong Screen, Markdown/HTML/JSON reports, and optional deep-analysis consensus o…

模型作用：Claude 4.5 Sonnet 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作：Claude Sonnet 4.5 is implemented as the default Anthropic model in the analyzer and is called to read resume text, optionally evaluate visual design from a resume image, apply the six-pillar rubric, return JSON scoring,…

A 类理由：有具体使用者、具体任务、公开原始证据和可访问产物，可公开核验。

风险边界：当前判断基于公开材料；若产物下线、仓库变更或模型参与比例仅来自作者自述，需要在引用前重新复核。 原始资料写作：Public open-source repository and README verify the task, artifact, and exact Anthropic model id; this is an independent tool rather than an Anthropic customer story, and users must supply their own Anthropic API key wi…

原始记录：AI PM Resume Analyzer uses Claude Sonnet 4.5 to score resumes against a six-pillar AI PM hiring framework

Mohamed Sorour 公开的代码代理与软件工程案例，来源为 公开代码库，复核于 2026-06-27T08:54:40Z。 Model Atlas 将它标记为 A 类证据，因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。

任务：这是一个围绕软件工程任务执行的真实任务，公开材料可以回溯到具体使用者和具体产物。 原始资料写作：Create a natural-language AWS operations assistant that monitors and optimizes cloud resources through Amazon Bedrock AgentCore Gateway, exposing CloudWatch, Logs, and EBS APIs as tools instead of requiring manual dashb…

公开产物：公开材料提供公开代码、README 或项目配置，可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作：The public repository contains a runnable agent, screenshots, setup instructions, local CLI flow, semantic tool discovery, and documented access to 137 AWS tools for monitoring metrics, analyzing logs, and optimizing EB…

模型作用：Claude 4.5 Sonnet 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作：The README names Claude Sonnet 4.5 as the LLM responsible for natural-language understanding inside the AgentCore Gateway plus Strands Agents architecture.

A 类理由：有具体使用者、具体任务、公开原始证据和可访问产物，可公开核验。

风险边界：当前判断基于公开材料；若产物下线、仓库变更或模型参与比例仅来自作者自述，需要在引用前重新复核。 原始资料写作：The README describes the project as production-ready while AgentCore Runtime deployment is still marked in progress; accepted as a public implementation artifact with clear model binding and task evidence.

原始记录：Mohamed Sorour built an AWS Resource Optimizer Agent with Claude Sonnet 4.5 and Bedrock AgentCore

asimniazi63 公开的代码代理与软件工程案例，来源为 公开代码库，复核于 2026-06-27T08:54:40Z。 Model Atlas 将它标记为 A 类证据，因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。

任务：这是一个围绕软件工程任务执行的真实任务，公开材料可以回溯到具体使用者和具体产物。 原始资料写作：Build an autonomous Enhanced Due Diligence research system that gathers intelligence on people and entities, analyzes risks, maps relationships, and generates comprehensive investigation outputs through a LangGraph-base…

公开产物：公开材料提供公开代码、README 或项目配置，可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作：The repository publishes the DeepAgents codebase and architecture for a production-ready due-diligence research agent, including orchestrator, Claude service, search service, result aggregation, error handling, and repo…

模型作用：Claude 4.5 Sonnet 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作：The README explicitly lists Claude Sonnet 4.5 in the multi-model architecture and assigns it to strategic analysis, query generation, and reflection steps in the due-diligence workflow.

A 类理由：有具体使用者、具体任务、公开原始证据和可访问产物，可公开核验。

风险边界：当前判断基于公开材料；若产物下线、仓库变更或模型参与比例仅来自作者自述，需要在引用前重新复核。 原始资料写作：The system also uses GPT-4o, so the case is multi-model; the Claude Sonnet 4.5 contribution is separately identified in the README and tied to specific workflow responsibilities.

原始记录：DeepAgents uses Claude Sonnet 4.5 for autonomous enhanced due-diligence investigations

Mark2009 公开的代码代理与软件工程案例，来源为 公开代码库，复核于 2026-06-27T08:54:40Z。 Model Atlas 将它标记为 A 类证据，因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。

任务：这是一个围绕软件工程任务执行的真实任务，公开材料可以回溯到具体使用者和具体产物。 原始资料写作：Port the jsonrepair JavaScript library to C++ and add repair capabilities for malformed JSON documents, producing a usable C++ implementation with examples.

公开产物：公开材料提供公开代码、README 或项目配置，可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作：The public repository contains the C++ jsonrepaircpp library, feature list, examples showing invalid JSON repaired into valid JSON, and attribution that the port was produced by Claude Sonnet 4.5 with additional repair …

模型作用：Claude 4.5 Sonnet 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作：The README states that the C++ port was created by Claude Sonnet 4.5, binding the model to the software translation and implementation work.

A 类理由：有具体使用者、具体任务、公开原始证据和可访问产物，可公开核验。

风险边界：当前判断基于公开材料；若产物下线、仓库变更或模型参与比例仅来自作者自述，需要在引用前重新复核。 原始资料写作：Evidence is the project README and public repository; no independent production deployment is claimed, but the artifact is a reachable code library rather than a tutorial, benchmark, or collection page.

原始记录：Mark2009 used Claude Sonnet 4.5 to port jsonrepair from JavaScript to C++

Max Harar 公开的代码代理与软件工程案例，来源为 公开代码库，复核于 2026-06-27T08:54:40Z。 Model Atlas 将它标记为 A 类证据，因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。

任务：这是一个围绕软件工程任务执行的真实任务，公开材料可以回溯到具体使用者和具体产物。 原始资料写作：Build an interactive airline customer-service agent against Sierra's open tau-bench airline domain, using the policy document as the system prompt, seeded airline data as state, and typed tools for tasks such as booking…

公开产物：公开材料提供公开代码、README 或项目配置，可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作：The repository publishes a working single-page live demo at maxharar.com/sierra and documents five scenario chips plus a Pass^1 measurement for claude-sonnet-4-5 on the selected airline tasks.

模型作用：Claude 4.5 Sonnet 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作：Claude Sonnet 4.5 is listed as the reasoning model in the stack and is described as driving the agent's policy-aware decisions and tool use over the airline-domain workflow.

A 类理由：有具体使用者、具体任务、公开原始证据和可访问产物，可公开核验。

风险边界：当前判断基于公开材料；若产物下线、仓库变更或模型参与比例仅来自作者自述，需要在引用前重新复核。 原始资料写作：The domain is based on an open benchmark environment, but the submitted artifact is a concrete public demo/repository rather than a leaderboard-only entry; repo and live demo were reachable during collection.

原始记录：Max Harar built a live Sierra tau-bench airline customer-service agent with Claude Sonnet 4.5