Compose the perfect prompt for the agent you already use

Investigate bug reports, diagnose root cause, implement fix with tests

Comprehensive code review with security scan and test verification

Fast code review of recent changes with quality and security checks

Scaffold a new project with structure, configs, and initial code

Analyze code quality, identify improvements, refactor with tests

Analyze coverage gaps and generate comprehensive tests

Generate unit and integration tests for existing code

OWASP Top 10 scan, dependency audit, and vulnerability report

Check for vulnerable, outdated, or unused dependencies

Auto-generate API docs, README, and inline documentation

Profile code, identify bottlenecks, suggest optimizations

Version bump, changelog, build verification, tag creation

Design, implement, and test API integrations with external services

Plan and generate database migrations with rollback support

Generate CI/CD pipeline configs for GitHub Actions, GitLab CI, and more

Deep dive into any codebase — explain architecture, patterns, and flow

Convert a PRD into a structured, ready-to-execute prompt for your coding agent

Turn a feature description into a PR-ready plan with file changes, tests, and rollout notes

Migrate JS or loose-TS to strict TypeScript file by file with type fixes

Add Stripe subscription billing with checkout, webhooks, and tier gating

Wire Supabase auth with email + OAuth, RLS policies, and session middleware

Add theme toggle with persistence, system preference, and no flash on load

WCAG 2.1 AA audit with prioritized fix plan

Add Sentry with source maps, release tracking, and performance monitoring

Find N+1 patterns in your ORM and propose batched/joined fixes

Find unused exports, unreferenced files, and dead branches

Identify and diagnose flaky tests; propose deterministic fixes

Generate a CHANGELOG.md entry from git commits, grouped by type

Design and scaffold a new MCP tool with schema, handler, and tests

Design a RAG pipeline: chunking, embeddings, vector store, retrieval, reranking

Generate reviewer-friendly PR descriptions from a diff

Initialize a monorepo with workspaces, shared configs, and build pipeline

Hand off mid-task work between Cursor, Claude Code, Copilot — produces clean state file and tailored prompt

Generate the .well-known/mcp-server-card.json required for MCP v2.1 registry discoverability

Add token-by-token streaming with cancellation, error recovery, and progressive markdown

Scan dependencies for malicious packages, typosquatting patterns, and CI/CD secret exposure risk

Wire persistent cross-session memory into an existing AI agent using Mem0 — store, retrieve, and summarize

Instrument an AI application with OTel spans for LLM calls, agent steps, and tool invocations

Find cacheable prompt prefixes in your AI app and add provider caching to cut LLM costs 60-80%

Diagnose CI build failures with AI — root cause analysis, breaking change location, and a fix plan or patch

Audit your release workflow for branch-hijack and import-time payload risks that postinstall-only scanners miss

Diagnose why an autonomous coding agent run went wrong and produce a root-cause fix from its tool-call transcript

Install a pre-execution hook that blocks destructive shell, git, and cloud commands before your AI coding agent runs them

Add the MCP Tasks extension to an existing server for long-running ops and human-in-the-loop approvals

Wrap your AI SDK calls with automatic multi-provider fallback so an outage or rate limit on one provider does not take the feature down

Have an AI agent find, confirm real exploitability of, and patch vulnerabilities in your changed code

Audit and speed up the lint/typecheck/test loop your coding agent depends on to self-correct