Claude Skill Marketplace

Create or edit Jira tickets at TeleClinic — type, title, description, fields, parent. Writes ticket and comment content only: no status transitions, no multi-ticket audits.

Writes and rewrites patient-facing TeleClinic copy for a specific channel (app screen, push, email, content card, IAM, LP headline, Meta ad script) and returns at least two channel-appropriate variants plus a recommendation with the psychological lever, the target funnel event and an HWG/UWG check. Pulls real patient language from Metabase for indication-specific copy.

Shapes Claude's output to be direct and immediately actionable: action first, numbered steps, no preamble or recap. A TeleClinic-renamed, de-pathologized version of the popular open-source "i-have-adhd" skill, per the Inclusive Language Guide (a diagnosis should not be used as a metaphor/brand for a communication style).

Guides you through the full bug-reporting process for doctor-related issues at TeleClinic: gathers information via a form, checks the Jira BUB board for duplicates, creates the ticket, and notifies the responsible team via Slack. Only activate when someone actively wants to report a bug or create a Jira ticket — not for general questions or troubleshooting.

Pulls the weekly Doctor Support KPIs (CSAT, survey participation, ticket density, treatment volume, and comment highlights) from Metabase and writes a bilingual (German + English) analysis text ready to paste into the company-wide Weekly report.

Runs TeleClinic's automated experiment learning loop: searches Won/Lost experiments, synthesizes cross-cutting learnings, and generates HBPS hypothesis and PG iteration tickets with a single approval gate before anything is written.

Explains how code the user works with actually works — a codebase, a framework, a merge request, a single file — for people who are not experts in that language or stack, grounded line-by-line in the repo's real code instead of generic textbook examples.

Creates a TeleClinic HBPS growth hypothesis from experiment keys or a problem statement, through to a validated Jira ticket, with a single approval gate before anything is written.

Creates formal ISO 27001 risk register entries for Teleclinic's ISMS, formatted for Vanta.

Scores a TeleClinic Prompt Battle session (chat history + artifact) with a weighted, evidence-cited verdict and a battle report for neutral side-by-side ranking.

Scan the market to find new potential integration partners for TeleClinic and return a scored, prioritised shortlist ready for outreach and HubSpot upload.

Checks finished patient-facing TeleClinic marketing copy (ads, landing pages, CRM, SEO) against HWG/UWG and the KV Nordrhein cease-and-desist settlement, and returns a traffic-light verdict per line with a compliant alternative that keeps the punch, plus the exact mandatory disclaimer where one is required.

Sizes up an opportunity or decision and returns a two-tab spreadsheet (the case, one arithmetic step per row, plus an evidence tab) and a short written verdict naming what the answer hinges on.

OpenEvidence clone for clinical evidence lookup — designed for TeleClinic's employed doctors, usable by anyone at TeleClinic.

Research a company, app, or product for TeleClinic partner outreach, then write a ready-to-send email and a co-branded landing page HTML mockup.

Creates and validates tickets (Idea, Requirement, Bug, Tech Debt) on the PLOPS Jira Product Discovery board PLOPSHE, following the board's current templates and fields, and only writes to Jira after explicit approval.

Personal prompting-craft feedback on any chat session, scored on TeleClinic's shared craft anchors, with concrete coaching and rewritten prompts.

Reviews a Claude skill (SKILL.md + bundled resources) for craftsmanship quality against TeleClinic's 15 skill design principles. Returns a traffic-light rating (red / yellow / green), a per-principle scorecard, and the highest-leverage fixes.

For marketplace maintainers: run a structured quality review of a submitted skill and grant/revoke the quality badge (★ rating).

Autonomously monitor a Slack thread with an open action item, checking every 2 minutes and driving the conversation to completion until the goal is achieved or the other party opts out.

Checks internal work content for AI slop before or after publishing. Gives a blunt publish / fix-first / don't-publish call and a traffic-light read across five dimensions: readability, slop phrases, language clarity, author comprehension, and context sufficiency.

Submit a Claude skill to the TeleClinic Skill Marketplace without a GitHub account or a terminal: packages it into the plugin structure, validates it, checks for secrets and patient data, and opens the pull request through the submission connector.

Runs a white-box OWASP Top 10 2021 static security audit of the TeleClinic Video platform (video-backend, app.teleclinic.video, teleclinic-video) and renders a TÜVIT-style penetration test report as a PDF.

Self-review and sign off your own spike / technical-investigation / proof-of-concept BEFORE you share it with stakeholders, so the usual gaps never make it into the doc. Enforces mandatory checks on success metrics, technical performance and authentication/integration mechanics, with extra attention to regulated/healthcare contexts (GDPR, patient data, ISO 27001, EU AI Act, MDR). Reports on the draft; it does not edit it.

Working handbook for the patient-app-ios codebase: house code-style conventions distilled from review feedback, build/test workflows (snapshot tests, SwiftUI previews, running on device, Maestro E2E), localization via Lokalise + SwiftGen, RemoteConfig A/B experiments, Mixpanel/Braze tracking, and task playbooks for the app's core flows. Use it whenever doing iOS development in that repo.