Backend in Engineering

Create algorithmic art with p5.js using randomness and interactive parameters.

Create original posters and visual art in PNG and PDF formats.

Build, debug, and optimize Claude API applications with caching and model migration support.

Build protocol servers that connect language models to external APIs and services.

Build complex React artifacts with Tailwind CSS and shadcn/ui components.

Test local web apps with Playwright, verify UI behavior, and capture screenshots.

Open, edit, and create Excel and CSV files with formulas, formatting, and data cleaning.

The official Supabase skill: Auth, RLS, Edge Functions, Realtime, Storage, migrations, and the JS/SSR clients — straight from the Supabase team.

Supabase's official Postgres performance playbook: indexing, connection pooling, pagination, locking, and query tuning.

Analyze your codebase to create a trimmed ECC install plan with only the skills you actually need.

Learn REST API best practices for resource naming, versioning, pagination, and error handling.

Delegate coding tasks to multiple AI models and get the best result.

Configure cross-session user memory and multi-profile isolation for AI agents.

Delegate coding tasks to OpenHands, a model-agnostic agent via terminal.

Query wallet balances, token info, gas prices, and transactions across 8 blockchain networks.

Build and optimize Claude API applications with prompt caching and model migration.

Build Model Context Protocol servers to connect AI with external APIs and services.

Delegate code tasks to compressed subagents that save context space.

Turn any codebase or document folder into an interactive knowledge graph with community detection.

Add persistent memory to Claude agents across sessions via context injection.

Manage job data and career operations through customizable plugins.

Scan job listings from Apify actors and map results into your career tracking system.

Extract job URLs from Gmail and add them to your career pipeline.

Generate professional architecture and workflow diagrams as HTML files with dark/light themes and image export.

Analyze git diffs and pull requests to understand changes, affected components, and risks.

Visualize your codebase's knowledge graph in an interactive web dashboard.

Ask questions about your codebase using an AI-powered knowledge graph.

Compare your code to the latest npm version and list what's changed.

Discover a hidden easter egg about the Oh My OpenCode development plugin.

Turn code folders into interactive knowledge graphs with community detection and query tools.

Design stable, well-documented APIs and module boundaries that are hard to misuse.

Implement the visitor pattern for type-safe discriminated unions with multiple dispatch sites.

Verify actual demand exists before reviewing code that adds new surface area.

Analyze, index, and generate wikis for your codebase via command line.

Trace bugs and errors by mapping code execution flows and finding root causes.

Diagnose your AI project and create an optimal technical roadmap in Spanish.

Manage Sub2API accounts, codes, proxies, and backend settings via admin API.

Route prompts to Claude, Codex, Gemini, or other AI models and save results.

Remove bloat and repetition from AI-generated code while preserving behavior.

Create isolated Herdr sessions for safe testing and debugging without affecting your main session.

Run browser automation on AWS Bedrock's managed cloud browsers.

Run headless Chrome browser automation in Vercel's ephemeral microVMs.

Writes and validates OpenAPI 3.1 specs from code or design-first contracts.

Separates read and write models with CQRS patterns for scalable, event-sourced systems.

Reference for service boundaries, inter-service communication, and resilience patterns.

Scaffolds production FastAPI projects with async patterns, DI, and pytest setup.

Split complex tasks across multiple focused agents while keeping strategic control in the parent.

Build Feishu and Lark bot automations, documents, sheets, and approval workflows securely.

Design, build, and debug Model Context Protocol servers with stdio and HTTP/SSE transports.

Search live tech job listings across 50+ job boards and multiple countries.

Search live Danish job listings by location, category, and employment type.

Manage LLM provider credentials and integrations in Arize for evaluators.

Create annotation schemas and manage human review workflows in Arize.

Generate customized instruction files for AI coding agents like Copilot and Claude.

Analyze your codebase and recommend Claude Code automations to optimize your workflow.

Automatically fix GitHub bug issues and submit pull requests.

Auto-fix high-frequency Sentry errors by analyzing stack traces and submitting pull requests.

Reduce Vercel costs and improve performance using verified metrics and framework-specific recommendations.

Pack codebases into single files optimized for AI analysis and exploration.

Write commits that capture intent, decisions, and constraints alongside code changes.

Generate and iterate high-performing ad copy, headlines, and variations for any platform.

Analyze competitor URLs to generate detailed profiles with market data and insights.

Generate and optimize images for blogs, social media, product mockups, and brand assets.

Build flexible, maintainable React components using composition techniques.

Optimize React and Next.js code following Vercel's performance best practices.

Animate page and component transitions using React's native View Transition API.

Create and edit embedded skills in Crush's binary system.

Create and manage custom shell commands for Crush's embedded shell environment.

Configure Crush settings, providers, servers, and permissions via JSON.

Add and debug hooks to gate, approve, or modify tool calls in Crush.

Index and reference guide for Agent Memory's MCP tools and parameters.

Let your agent review code using its own intelligence while OCR handles file filtering.

Trace and debug code issues using semantic search and execution flow analysis.

Build and maintain a knowledge graph of your codebase for smarter reviews.

Take a quick quiz to find your starting point in an AI engineering curriculum.

Build knowledge graphs from documents and search them with natural language queries.

Search Slack chat history to find solutions, code snippets, and team decisions.

Analyze customer accounts to predict churn risk and identify expansion opportunities.

Generate SQL queries from plain English descriptions for BigQuery, PostgreSQL, MySQL, and more.

Build serverless apps, APIs, and infrastructure on Cloudflare's global edge network.

Answer questions about Effect, a TypeScript framework for robust applications.

Build stateful AI agents and workflows on Cloudflare Workers with real-time capabilities.

Triage bugs, find duplicates, and file GitHub issues with full context.

Format Linear or Superset tickets with a standardized three-section structure.

Debug queen sessions, colony forks, and worker spawns without affecting production.

Create and edit Obsidian Canvas boards with nodes, links, and visual layouts.

Triage bugs, find duplicates, and file GitHub issues with full context.

Create, discover, and run domain-specific skills bound to automated workflows.

Control Chrome via CDP to automate browser tasks and reuse login sessions.

Interact with OmniRoute A2A server, send tasks, and test the agent-to-agent protocol.

Build tools to import, inspect, and migrate DeepChat chat data and configurations.

Format command-line messages and help text with consistent colors and patterns.

Generate an AI design contract and evaluation strategy for system development phases.

Generate and evaluate improvement ideas for your project workflow.

Scan your entire codebase to find dead code, over-engineering, and simplification opportunities.

Display a quick-reference guide for Ponytail modes, skills, and commands.

Find over-engineered code, unnecessary dependencies, and complexity to delete.

Find the simplest solution that works, questioning if the task is necessary.

Generate flowcharts, architecture diagrams, and visual system maps as PNG, SVG, or PDF.

Query a codebase graph to understand code structure and dependencies without reading source files.

Convert markdown workflow specs into compiled GitHub Actions using AI engines.

Generate and maintain a source-grounded wiki documenting your codebase architecture and modules.

Deploy static files to IPFS or full-stack web projects with backend APIs and databases.

Extract text and visual details from images that Claude cannot see natively.

Add a new Roslyn diagnostic rule with metadata, code fix, and tests.

Compress AI context by up to 99% when reading files, searching code, and running commands.

Index, organize, and manage your codebase with search and code graph infrastructure.

Search and understand code structure, dependencies, and architecture using semantic analysis.

Autonomously research, benchmark, and implement performance optimizations with production-ready pull requests.

Measure and analyze Go application performance with benchmarks, profiling, and statistical comparisons.

Build and extend command-line tools in Go with proper structure and flag handling.

Design reliable, scalable data systems with proven patterns for consistency, replication, and streams.

Launch and monitor AI coding agents in tmux with session management and parallel worktree support.

Search and index your codebase semantically to find relevant code quickly.

Turn website data extraction into reusable, deployable skills with automatic API discovery.

Audit your CLAUDE.md file for vague instructions, stale facts, and unnecessary bloat.

Create, analyze, and modify Word, Excel, and PowerPoint files without Office installed.

Build custom extensions for Super Agent Party with frontend or Node.js backends.

Convert security findings into a defensible risk score and dollar-based breach cost estimate for board reports.

Generate standards-compliant DCAT feeds for your PortalJS portal to enable harvesting by open-data registries.

Create and register code patches for Claude Code customizations.

Convert any webpage to clean markdown with rendered HTML snapshot.

Avoid common GA4 API pitfalls that return misleading data instead of errors.

Generate audio narration of blog posts with multiple voices and podcast modes.

Build ChatGPT and MCP apps from concept to deployment.

Guide developers through creating and deploying ChatGPT apps with tools and custom UI.

Guide developers through creating and deploying MCP apps with tools and custom UI views.

Fix bugs quickly with open → build → verify → archive workflow.

Generate design documents and specifications through structured brainstorming.

Identify pitfalls in Google Search Console API queries to avoid silent errors.

Add a new OAuth provider to your registry with platform approval and connection flows.

Add performance profiling to your Rust project with zero overhead when disabled.

Capture diagnostic information and file bug reports for OpenKnowledge issues.

Maintain an agent-authored wiki documenting your codebase's architecture, modules, and flows.

Import media files into OpenChatCut projects for video editing.

Edit local video projects with timeline, captions, and motion graphics tools.

Create and update feature task plans with progress tracking and blocker management.

Set up and run OpenTag to connect collaboration platforms with local coding agents.

Generate multiple competing explanations for an observed problem or signal.

Decompose boundary statements into legal, admissibility, deontic, and evidence categories.

Add new AI assistant providers to ClaudeBar following test-driven development patterns.

Structure code so business logic stays independent from frameworks and databases.

Create generative art with p5.js using custom algorithms and interactive parameters.

Create complex React artifacts with Tailwind CSS and shadcn/ui components.

Create original visual designs as PNG and PDF files using design philosophy.

Build Model Context Protocol servers to connect LLMs with external APIs and services.

Detect AI-generated audio, images, video, and text to identify deepfakes and synthetic media.

Test local web applications with automated browser interactions and debugging.

Search, transcribe speech, and extract text from documents using Azure AI.

Fetch and summarize pull request review comments into actionable items.

Capture bug details and create structured reports your team can act on.

Design database schemas visually with entity-relationship diagrams and Prisma format.

Design and optimize SQL, NoSQL, and vector databases with schemas, indexing, and compliance.

Build and review APIs, databases, authentication, and server logic with clean architecture.

Analyze software architecture decisions with tradeoff analysis, risk assessment, and stakeholder synthesis.

Organize branch chains and manage stacked rebases, merges, and PR workflows.

Find, download, and track movies, TV shows, and anime from your local media agent.

Design schemas, optimize queries, and choose the right database for your application.

Build Node.js backend AI apps with text and image generation models.

Call AI models directly from browser apps using CloudBase JavaScript SDK.

Generate and stream text using AI models in WeChat Mini Programs.

Handle server-side authentication, user lookup, and custom login in Node.js CloudBase apps.

Search previous coding sessions for relevant context, decisions, and past attempts.

Transcribe audio and video files to text with fast local or remote processing.

Investigate and resolve Cloudflare configuration issues using API-driven evidence gathering.

Merge redundant documentation files while keeping all unique content and information.

Transform meeting transcripts into structured minutes with speaker identification and iterative review.

Convert markdown documents to professional PDFs with Chinese font support and theme options.

Orchestrate complex tasks with parallel verification agents and strategic execution gates.

Find new Japanese NLP GitHub repositories missing from awesome-japanese-nlp-resources lists.

Run Claude Code with external AI models through sub-agents.

Delegate coding tasks to Antigravity CLI and review diffs before landing changes.

Convert product requirements into structured epics with user stories and spec documents.

Extract YouTube metadata, subtitles, and download videos with proxy support.

Scrape websites and extract structured data without getting blocked.

Follow code quality, testing, and contribution guidelines for the Marmite codebase.

Convert markdown files into a complete static website with zero configuration.

Recreate award-winning creative websites with forensic precision by reverse-engineering their code.

Store and retrieve decisions, discoveries, and patterns across chat sessions.

Create and configure MATLAB experiments to sweep parameters and compare configurations.

Import PyTorch, ONNX, and Keras models into MATLAB as dlnetwork objects.

Automate website interactions like filling forms, clicking buttons, and scraping data.

Classify and resolve conflicts when ingesting data bundles.

Start or restart a local Temps development server with isolated ports per branch.

Access stock quotes, charts, and financial data via the Yahoo Finance library.

Control and inspect live browser panels within Horizon workspace.

Find the right AI model for your task with capabilities and cost comparisons.

Query and manage MySQL or PostgreSQL databases on AWS Aurora with serverless scaling.

List, monitor, inspect, and control active Bernstein agents in real-time.

Choose the right development approach for your project based on current state analysis.

Analyze any codebase and explain what it does, why, and what alternatives exist.

Convert conversations into Flow Next spec files with review before saving.

Train custom anime character and style adapters for the Anima model locally.

Run questions across multiple AI providers and synthesize their responses.

Delegate coding tasks to a separate Claude session, review the diff, and merge the changes.

Convert approved game design and art direction into a playable game build.

Scaffold new external service integrations with credentials, auth polling, and settings UI.

Generate changelogs automatically from Git commit history.

Identify performance bottlenecks and generate optimization suggestions for code, databases, and networks.

Generate CI/CD workflows for GitHub Actions based on your project's tech stack.

Analyze GitHub issues for validity, priority, and recommended actions.

Investigate Sentry issues, events, traces, and logs from the command line securely.

Resume conversations later when background tasks like CI jobs or reviews complete.

Build a mental map of your codebase once, then answer questions across multiple turns.

Calculate your agent's productivity score from session metrics and cache efficiency data.

List, query, version, and clean up model data artifacts with flexible filtering.

Generate a checklist document of Java system prompts from your skills inventory.

Generate a Java agents inventory checklist document from a predefined template.

Generate professional system architecture diagrams as shareable HTML and SVG files.

Write, edit, and debug Vera code, a functional language with typed slots and mandatory contracts.

Convert a photo of handwriting into an installable font file.

Run security, quality, and performance audits on entire codebases with detailed, quote-backed reports.

Draft replies to code review questions using implementation reasoning.

Find routes, locations, distances, and elevation data for any geographic query.

Learn the architecture and design decisions for the Google Maps MCP server project.

Get code review, debugging help, and design feedback from Claude via CLI.

Consult Gemini for coding tasks like prototyping, debugging, and code review.

Discover and extract all URLs from any website's sitemaps.

Show available harness templates and what each scaffolding option includes.

Register and manage agent accounts, subscriptions, and API keys.

Deploy and manage generative AI models and ML workflows on SAP's enterprise cloud platform.

Check if your local kernel matches the harness scaffold version.

Scaffold an AI agent harness in one command from pre-built templates.

Scaffold a custom AI agent harness with CLI, MCP server, and marketplace identity.

Generate interactive architecture diagrams from codebases or system descriptions.

Create and manage Jira tickets from requirements and requests.

Build modular data processing workflows with batch jobs and API integration.

Find, audit, and design AI agent loops with safe triggers, actions, and stopping conditions.

Create and configure AI agents for automation, coding, research, and support tasks.

Train and optimize AI agents using reinforcement learning and automatic prompt tuning.

Build self-improving AI agents using 9 reinforcement learning algorithms.

Store and search AI agent memories locally with mathematical ranking and entity graph scoring.

Learn patterns from past tasks and share them across multiple agent loops.

Create and manage database schema changes with Flyway naming conventions and versioning.

Monitor Claude Code sessions and approve tasks remotely via Telegram, Feishu, or web dashboard.

Read, edit, and create Microsoft Word documents with tracked changes and comments.

Migrate embedding models in Qdrant without downtime or data loss.

Generate, analyze, and debug Verilog RTL designs with synthesis validation.

Design flexible, reusable content schemas for headless CMSes.

Analyze code architecture, call graphs, and refactoring issues without AI.

Transform feature ideas into annotated stories with automated greenfield or brownfield detection.

Train custom small language models using knowledge distillation and production data.

Debug code issues by forming hypotheses from structural knowledge before reading code.

Apply a confirmed refactoring plan to your codebase step by step.

Generate Chen-model ER diagrams from SQL or DBML schemas with export options.

Generate HTML reports showing your Claude Code usage, projects, and activity statistics.

Test trading strategy parameters and generate performance heatmaps.

Run multi-turn AI analysis tasks with database and RAG context.

Generate a complete backtesting script for trading strategies with data, signals, and performance charts.

Parse, build, and serialize JSON with SnowBank's high-performance DOM and code generators.

Create or migrate functional skills from requirements or legacy code.

Normalize skill requests into structured specifications for skill generation.

Analyze legacy skills to plan refactoring into modular components.

Guide a greenfield project through structured architecture planning before coding.

Compress and optimize 3D Gaussian Splatting models for web, mobile, and edge deployment.

Query scientific figures, select visual examples, and export reference bundles for plotting.

Evaluate and fix unresolved pull request review comments from any reviewer.

Automate browser tasks like scraping, form-filling, and screenshots on JavaScript-heavy sites.

Characterize and size MOSFETs using gm/ID methodology with ngspice simulations.

Search local files and content with progressive context levels to minimize token usage.

Interview you to create project specs, charters, and acceptance criteria.

Brainstorm ideas with a partner LLM in alternating turns, then synthesize the best into a plan.

Break complex work into atomic tasks executed in fresh context windows.

Decide whether an AI agent should run code or reason in text for each task.

Stream real-time and historical US stock data from Alpaca via REST and WebSocket.

Prepare a winning hackathon submission with track selection, demo scripts, and judging checklists.

Build, debug, and deploy Salesforce agents using Agent Script.

Review, debug, and mentor Python and JavaScript code with security and testing rigor.

Query wallet balances, token portfolios, transactions, and NFTs on Solana with USD pricing.

Read, create, and manage notes in Obsidian vaults from the command line.

Monitor Treasury and bank liquidity impacts during US government shutdowns.

Extract keywords and patterns from converted text recordings.

Record and query project decisions with timestamps, rationale, and consequences.

Generate automation scripts for browser agents to bundle multi-step interactions efficiently.

Track marketing funnel metrics from social posts through bookings.

Discover and test APIs for security vulnerabilities like BOLA, JWT attacks, and business logic flaws.

Map attack chains and kill paths from reconnaissance data using MITRE ATT&CK.

Map attack surface with subdomain enumeration, port scanning, and technology fingerprinting.

Connect to external services through managed API routes.