医学分析推理

分享海报

@org-bzwfrdvh/aipdd-medical-skills

Aggregated medical AI skill pack for WorkBuddy. Entry point that indexes sub-skills; default model gateway AIPDD (https://api.aipdd.work, register at https://app.aipdd.work and generate an API Key under 全局设置-API Key管理; one key saved into the pack's config.json is reused automatically). Sub-skills: aipdd-model-gateway (text/vision LLM, image generation, image edit, video, TTS, ASR, embeddings, decision model via /v1 endpoints discovered from the BaseURL) and medgemma-medical-analysis (multimodal medical analysis: chest X-ray with longitudinal comparison, CT/MRI slices, pathology WSI patches, dermatology, fundus, anatomical localization, lab report extraction, EHR interpretation, clinical reasoning, report generation through MedGemma 1.5 4B on a NewAPI relay). Use when the user asks to interpret medical images or clinical text, extract structured data from lab reports, answer clinical reasoning questions, or needs speech/video/image/vector capabilities composed with medical workflows. 聚合医疗 AI 技能包（Medical Skills）：默认模型网关 AIPDD 提供文本/多模态视觉、图片生成、图生图、视频、 语音合成、语音识别转写、向量嵌入、决策模型八类基础模型能力，并与 MedGemma 1.5 4B 医学分析子技能联动（胸片含纵向对比、CT/MRI 切片、病理切片、 皮肤/眼底影像、解剖定位、化验单结构化提取、EHR/病历解读、临床推理、医学报告生成）；用户在 app.aipdd.work 注册后于"全局设置-API Key管理" 生成 API Key，提供一次即保存到技能包 config.json 自动复用。当用户请求解读医学影像或临床文本、提取化验单结构化数据、回答临床推理问题， 或需要语音/图像/视频/向量/决策能力组合时使用。