10.8K skills · sorted by relevance
Draft personalized sales emails, call scripts, and LinkedIn messages using Common Room activity data and recent public news. Each draft connects a specific prospect signal to the outreach, with notes explaining the chosen angle and any assumptions.
Plan large sets of search-focused pages using templates and data, such as location, comparison, and integration pages. Define unique content, URL patterns, internal links, launch checks, and monitoring metrics for a structured publishing plan.
Build or update Figma screens from code or descriptions using an existing design system, a shared set of components and visual rules. Assemble sections with reusable components, colors, spacing, and styles, then check their appearance.
Write, revise, and review beads user documentation using consistent terminology, explanations, diagrams, and navigation. Follow project rules for generated pages and release alignment, with checks for links, freshness, and documentation consistency.
Prioritize product or business assumptions using an impact and risk matrix. Sort them into implementation, deferral, rejection, or testing, then propose experiments for assumptions that need validation, using actual behavior and clear success metrics and thresholds.
Develop applications with the Claude API using examples from Anthropic's official cookbooks. Explore document summarization, image analysis, document retrieval, external tool calls, and prompt caching through practical Python code and implementation guidance.
Evaluate Google Cloud workloads against built-in best practices or custom organizational rules using Workload Manager. Use public client libraries or the REST API to run checks, review violations and scanned resources, and export results to BigQuery.
Create and validate Terraform alerts for GKE using PromQL, a monitoring query language. Cover response times, errors, traffic, memory usage, and cluster health, with checks for alert timing, duplicate signals, and metric ingestion costs.
Assess how changes, failures, or paused updates to a BigQuery table or view affect downstream assets. Trace data lineage, the relationships showing where data flows, and summarize dependent tables, dashboards, and processes with direct and indirect impacts.
Find content losing search impressions and clicks using Google Search Console. Rotate through pages based on refresh history, update dated copy, statistics, and internal links, and open a pull request for human review.
Investigate a codebase, reproduce bugs, implement requested changes, and verify the results using Leon’s file, search, and shell tools. Follow project instructions, preserve existing user changes, and report checks and unresolved issues.
Guide language model training with TRL, from learning to follow instructions to matching human preferences. Covers method selection, training data, reward models that score responses, evaluation, and adjustments for memory limits or unstable training.
Build Vue 3 applications using JavaScript and JSDoc comments that describe data types. Create components and reusable logic, configure Vite, routing, and state management, and check annotations and behavior with ESLint and Vitest.
Handle order refunds by checking the supplied refund policy first, calling refund-order with an eligible order’s ID, and using the provided confirmation template to respond to the customer.
Diagnose paid advertising performance using account evidence, financial constraints, and experiments. Draft ranked changes to budgets, bids, or ad delivery, and apply supported operations only after approval and state checks, retaining a way to reverse them.
Generate talking-person videos from a portrait or an existing video and an audio track, aligning lip movements, expressions, and body motion with speech. It supports long-form generation and can create speech from text using Kokoro-82M.
Run repeated code reviews and fixes using an installed reviewer chosen by the user. The workflow tracks findings across rounds, checks that edits stay within scope, and reports whether blocking issues were resolved or further decisions are needed.
Assess customer acquisition channels using acquisition costs, lifetime customer value, payback time, retention, and growth capacity. Compare channels and receive recommendations to increase spending, improve performance, pause activity, or continue with a spending cap.
Check article and file availability through the NCBI PMC Open Access service using supported article identifiers. Receive a concise metadata summary, with the option to save the original XML response when you need the full service output.
Create original generative art with p5.js, using code to produce patterns and motion. Receive a written artistic concept and an interactive HTML viewer with adjustable parameters, reproducible random variations, seed navigation, and PNG downloads.
Apply related edits across several files using a unified diff, a file describing code changes. Check the patch with Git before applying it, and rebuild rejected patches using the current file contents.
Review existing tests using 19 research-backed test smell categories, meaning patterns that can weaken test quality. Get findings ranked by demonstrated risk, with precise code evidence and fixes suited to the test framework in use.
Turns a copy-edit report into conflict markers that display original and suggested text together in the source file. Review each proposed change using VS Code or Positron's merge interface, optionally limiting edits to one Quarto section.
Analyze whether an intervention causes an outcome using DoWhy in Python. Build a graph of causal assumptions, determine whether the effect can be estimated, compare estimates, and test how well results withstand changes to assumptions or data.