We have seen countless AI systems that can write poems and hold conversations, but far fewer medical “agents” that can truly understand clinical records, make sense of diagnostic and treatment reasoning, or even assist with medical insurance reviews.

A general-purpose foundation model may accurately recognize every character on the page, yet still struggle to penetrate the medical logic behind the text. It can generate fluent prose, but it cannot reliably determine whether a diagnosis conflicts with a medication plan, whether the evidence chain is complete, or whether a case complies with medical insurance review rules. In serious healthcare settings, the real threshold has never been simply “being able to read and write.” It is whether AI can understand, guide, collaborate, and take responsibility like an experienced expert.

From an “intelligent tool” to a “medical expert,” this is not only a technological evolution, but also a redefinition of the role AI can play.

Starting today, Unisound’s five core medical Skills are officially available in the ClawHub community. They cover five capabilities: medical knowledge Q&A, outpatient medical record generation (initial visit/follow-up), critical illness review, and chronic disease review. This is more than opening up technical capabilities. It packages a “living medical logic engine” into intelligent entities that can be called on demand, giving AI the reasoning core of a “medical expert” and enabling precise services for hospitals, insurance institutions, and individual users. These capabilities can also be combined to meet the compound needs of patients, family members, physicians, insurance institutions, insurance reviewers, and other roles.

01 The Real Bottleneck in Medical AI: Not “Parameters,” but “Logic”

Medical AI has moved beyond the early stage of basic recognition. Speech transcription and information extraction have become standard capabilities. Yet why does AI still struggle to earn full trust in critical areas such as diagnosis and treatment, and medical insurance review?

Because between “being able to read a report” and “understanding clinical practice” lies a professional barrier of logic.

A critical illness review dossier often contains heterogeneous data from multiple sources, including outpatient records, discharge summaries, pathology reports, and surgical records. A well-formed medical record, meanwhile, contains rigorous causal relationships among the chief complaint, history of present illness, and past medical history.

A general-purpose model may read every page of text, but it cannot necessarily answer the questions that matter: Does the diagnostic classification conflict with the medication plan? Do the timing of key examinations and tests comply with clinical standards? Are there gaps or contradictions in the medical evidence chain? Are the progress notes logically coherent and compliant? These core questions—rooted in clinical reasoning and medical insurance rules—are what healthcare scenarioses truly require from AI.

Unisound’s five medical Skills now available on ClawHub draw on years of deep experience in healthcare, together with extensive accumulated clinical data and medical knowledge, to bridge the final mile from “text processing” to “clinical logic understanding.” More importantly, the backend knowledge supporting these Skills is continuously updated: as medical guidelines evolve and medical insurance policies change, the underlying “medical logic entities” evolve in sync, helping ensure that AI remains professionally up to date.

02 Five Skills: Each One a “Medical Expert” on Your Team

Unlike the basic functions of general-purpose AI, Unisound’s five medical Skills are designed entirely around real clinical workflows. Each capability corresponds to a specific type of “medical expert” assistant, transforming AI from a “passive tool” into an “active collaborator.” The Skills do not operate in isolation. Built on medical knowledge bases and insurance rule systems, they can work together to provide cross-scenario services for hospitals, insurers, and individual users.

(1) Medical Knowledge Q&A Skill: A “Health Advisor” for Patients

Direct Skill link: http://clawhub.ai/yishuihan132/medical-qa

• Core Capability: For patients who want to perform basic health self-checks, the Skill helps them understand symptoms, interpret examination and test results, learn general medication knowledge, and receive professional guidance on when to seek medical care. Based on authoritative medical knowledge bases—including drug labels, clinical guidelines, and reference ranges for test items—the system provides friendly, conversational health Q&A in a warm and empathetic tone.
• Value: Helps users better understand symptom context, medication information, and disease progression; proactively guides them to seek medical care when appropriate; and does not provide prescription recommendations, keeping safety boundaries clear.
• Logic Core: Integrates authoritative medical knowledge, explains complex medical terminology, and helps bridge the gap between the public and professional healthcare.

• Core Capability: Accepts noisy, unstructured transcripts of doctor-patient conversations; intelligently extracts key diagnostic and treatment information in accordance with clinical documentation standards; and automatically generates compliant, structured outpatient medical records, including the chief complaint, history of present illness, past medical history, and more.
• Value: By the time the consultation ends, a draft medical record is ready as well. Physicians only need to review and confirm it, freeing them from tedious manual data entry and allowing more time for clinical care itself.
• Logic Core: Embeds the medical logic of clinical record writing to ensure causal consistency between the chief complaint and the history of present illness.

• Core Capability: Reads unstructured historical follow-up records, automatically identifies and extracts key information—such as changes in chief complaints, medication adjustments, and test results—and converts it into structured fields so physicians can quickly review a patient’s history.
• Value: Helps physicians quickly understand the evolution of a patient’s medical history during follow-up visits, avoiding repeated review of lengthy documents. It also provides high-quality structured data for clinical research and data governance.
• Logic Core: Based on patterns of disease-history evolution, it accurately extracts the core fields relevant to outpatient records. It complements the Initial-Visit Medical Record Generation Skill to support full-process management of outpatient medical records.

• Core Capability: Processes heterogeneous medical materials from multiple sources—including medical records, imaging, and surgical records—and, based on critical illness insurance terms and medical insurance review rules, automatically extracts key information and aligns supporting evidence. It can identify inconsistencies across documents and missing evidence.
• Value: First understand the materials and organize the evidence, then leave the final decision to people. This enables a leap in efficiency from “manual document review” to “intelligent pre-review,” with traceable review conclusions.
• Logic Core: Embeds critical illness claims logic and validates the evidence chain across diagnoses, surgeries, examinations, and tests.

(5) Chronic Disease Review Skill: The “Eligibility Expert” for Qualification Assessment

Direct Skill link: http://clawhub.ai/aaiccee/chronic-disease-review

Core Capability: For a target chronic disease category, automatically extracts the key evidence elements required for disease qualification, then performs rule-based checks and consistency validation.

• Value: For common chronic diseases such as diabetes and hypertension, accurately identifies diagnostic evidence and medication use. This reduces repetitive review and improves both review efficiency and consistency of results.
• Logic Core: Based on chronic disease management guidelines and medical insurance eligibility rules to ensure standardized reviews.

03 Collaborative Workflows: Continuously Evolving Professional Knowledge to Unlock Cross-Skill Value

All five Skills are backed by continuously evolving medical knowledge and medical insurance rules. As clinical guidelines are updated, medical insurance policies change, and new drugs enter the market, the underlying “medical logic entities” are iterated in sync, helping keep every Skill at the forefront of professional knowledge.

More importantly, the Skills can be combined across use cases to support more complex needs:

• Individual physicians (e.g., primary care general practitioners): Individual physicians (e.g., primary care general practitioners): use the Initial-Visit Medical Record Generation Skill to support routine consultations → use the Follow-Up Medical Record Generation Skill to organize patient history and quickly review prior conditions → use the Critical Illness Review Skill to quickly assess whether a patient meets a specific critical illness criterion → use the Medical Knowledge Q&A Skill to explain the condition to the patient in plain language.
• Data exchange between insurance institutions and hospitals: When an insurer calls a review Skill and detects a gap in the evidence chain, it can precisely identify the additional materials required. Hospitals can use the medical record generation Skills to standardize documentation and improve data quality at the source, indirectly improving the efficiency of downstream claims processing.
• Clinical research and data governance: The Follow-Up Medical Record Generation Skill can process historical medical records in batches. The extracted standardized fields can be imported directly into analytics tools, shortening data governance work that once took months to a matter of days and accelerating research output.
• Closed-loop personal health management: After using the Medical Knowledge Q&A Skill for self-checking, a patient who needs medical care can authorize the Q&A record to be imported into a hospital system. The receiving physician can then use a medical record generation Skill to quickly capture key information and conduct the consultation more efficiently.

04 Redefining Medical AI: From Functional Tool to Core Clinical Collaborator

What determines whether medical AI earns a lasting place is not its accuracy in a single test, but the professional trust and practical value it builds over long-term use.

When AI is no longer merely a tool for processing text, but an agent capable of understanding clinical logic, validating evidence chains, and producing traceable professional conclusions, it completes the transformation from a “function call” into a “medical expert.”

The launch of Unisound’s five medical Skills on ClawHub marks a new step toward open access to specialized medical AI capabilities. Unisound is lowering technical barriers by opening years of accumulated “AI-powered clinical reasoning capabilities” to developers across the industry in the form of lightweight Skill entities. More importantly, the medical logic and rule engines behind these Skills will continue to be updated and evolve alongside the industry.

Starting today, developers across the industry, hospitals, insurance institutions, health management professionals, and individual users can access Unisound’s dedicated medical Skill pages through the ClawHub website and invoke core medical AI capabilities—including medical record generation, medical insurance review, and knowledge Q&A—with a single click. Make AI a true “medical expert” colleague in your workflow.