Prompts
Prompts
A curated library of ready-to-use prompts for AI models, agents, and automation workflows. Copy, tweak, and ship.
Transform raw breaking changes, diffs, and notes into a structured, step-by-step migration guide for consuming developers.
Define a state schema, persistence pattern, and resume logic for long-running automations to ensure safe restarts, idempotency, and data integrity.
Generate a structured testing plan and scoring rubric for validating AI agent behavior, catching regressions, and ensuring output quality.
Draft reversible migration script with pre-checks, rollback, and verification steps for your target database engine.
Generate system prompts that instruct upstream AI agents to package state, data, and metadata for seamless handoffs in multi-agent workflows.
Evaluate short-term context retention, vector storage, and state persistence patterns for your custom AI agent or automation pipeline.
Generate structured, developer-ready API reference documentation from raw controller code, route definitions, and type signatures.
Generate mock payloads and a structured test plan for edge cases, failures, and success paths in your automation workflow.
Architect how an AI agent stores, truncates, summarizes, and retrieves working state and long-term memory across turns.
Refactor untestable code by isolating side-effects with dependency injection and generating isolated unit tests.
Generate production-ready system prompts for AI agents using external tools, with explicit tool-choice logic and error fallbacks.
Deconstruct an academic paper, RFC, or technical spec into a production implementation plan, data flow architecture, and bottleneck analysis.
Convert fragmented feature notes and tickets into a structured engineering RFC for team review.
Architect a robust queueing and rate-limiting strategy for automations subject to third-party API rate limits and concurrency caps.
Create structured handoff contracts, context truncation rules, and state schemas for delegating tasks between AI agents.
Transform fragile, one-off scripts into safe, idempotent tasks with explicit state checks, failure handling, and execution safeguards.
Refactor complex prompts built for frontier models so they run reliably on smaller, cheaper models like Haiku or GPT-4o-mini.
Stress-test a proposed code snippet, config, or workflow against hidden risks, race conditions, and failure modes before you commit.
Synthesize code diffs, commits, and ticket context into a structured PR description with testing strategy and reviewer focus flags.
Architect an idempotency, payload validation, and deduplication strategy to prevent double-processing in webhook-driven workflows.
Create fault-tolerant workflows with automatic recovery strategies, state snapshots, and escalation triggers for multi-step agent processes.
Analyze package manager errors and constraints to propose safe resolution strategies with risk assessment.
Audit system prompts to find direct contradictions, ambiguous constraints, and rule priority collisions before deploying to production.
Systematically research an open-source library for security, license, maintenance, and performance risks before adopting it in production.
Convert unstructured tech discussions, PR comments, or Slack threads into a standardized Architectural Decision Record (ADR).
Generate a robust error-handling, rollback, and retry strategy for any automated workflow to prevent silent failures and data corruption.
Define precise programmatic triggers, risk categories, and state-resume strategies to keep autonomous agents safely supervised.
Create a defensive wrapper with retries, timeouts, and fallbacks for an unreliable external API or service.
Generate a production-ready LLM evaluator prompt with strict scoring rubrics and bias mitigations to benchmark AI agent outputs.
Research and audit a third-party API's constraints, rate limits, and webhooks before you write any integration code.
Transform raw git commit messages or pull request titles into a structured, reader-friendly changelog categorized by impact.
Generate clean Python or JavaScript code to transform and map complex webhook payloads between mismatched APIs in Make, n8n, or Zapier.
Generate high-quality, diverse few-shot examples (including edge cases and error handling) to insert into your system prompts for better accuracy.
Generate a structured research plan for a bug/error. Ranks root cause hypotheses, suggests verification steps, and flags missing info instead of guessing.
Audit an existing automation workflow to identify single points of failure, missing error handling, rate limit risks, and idempotency issues.
Extract a minimal, concrete reproduction from a messy user-written bug report before you spend time trying to reproduce it yourself.
Get a structured, decision-ready comparison of two implementation options instead of a wishy-washy 'it depends' answer.
Write a tool/function definition an LLM agent will actually call correctly — the part most tool-use bugs trace back to.
Turn a raw incident timeline into a structured, blameless postmortem with concrete follow-up actions instead of vague promises to 'do better.'
Turn a use case into a concrete Model Context Protocol server design — which tools, which resources, and where the boundary between them should sit.
Extract clear owners, deadlines, and decisions from rough meeting notes without losing the details that matter.
Have a model attack-test your agent's system prompt for prompt injection and instruction-override risks before real users do.
Turn a raw error and stack trace into a ranked list of likely root causes instead of a guess-and-check debugging session.
Turn a feature description into a concrete test plan that actually covers edge cases, not just the happy path.
Turn a plain-language automation idea into a concrete trigger-action blueprint for Zapier or Make, including where it will break.
Generate a production-ready system prompt that enforces strict JSON/schema compliance, negative constraints, and few-shot examples for reliable automation.
Systematically figure out why a retrieval-augmented generation pipeline returned a wrong or irrelevant answer.
Cut a technical draft down to its essentials without losing precision or turning it into marketing fluff.
Turn a plain-language automation idea into a concrete, node-by-node n8n workflow plan.
A reusable skeleton for writing a system prompt for a tool-using AI agent — role, tools, guardrails, and output format.
Get a clear, jargon-aware explanation of a technical concept, pitched at exactly the right level.
Have an AI model review a diff for correctness, security, and maintainability issues before you open a PR.