Design an AI Agent With Tools, Guardrails and Eval Cases Before You Build It
Turns an agent idea into a build-ready design: the job, the tools with their inputs and permissions, guardrails, failure handling and a set of evaluation cases to test it.
341 prompts · last added
79 results for “except for one eye” · page 2 of 4
Turns an agent idea into a build-ready design: the job, the tools with their inputs and permissions, guardrails, failure handling and a set of evaluation cases to test it.
Produces an onboarding guide for one folder of a codebase: what it does, how data moves through it, the files to read first and the traps a new developer should avoid.
Writes a verified handoff file covering changes, decisions, open issues and next steps so a new Claude Code session can continue without re-reading the whole history.
Has Claude Code review a branch diff like a skeptical maintainer, producing severity-ranked findings with evidence and a clear merge, fix, or block recommendation. Use it before opening or approving a pull request.
Audits an existing multi-agent pipeline for failure modes and produces retry limits, fallback paths, escalation triggers, and a test plan. Supports the decision on what to harden before wider rollout.
Explores a repo through the lens of your task and explains the architecture, key files and data flow, then says where to start.
Has Cursor lock in current behavior with characterization tests before restructuring a legacy module in small steps, so you can decide whether the refactor preserved behavior and is safe to merge.
Turns an error report, stack trace, and logs into ranked root-cause hypotheses. Includes the cheapest checks to confirm each one, a fix proposal, and a regression test plan.
Turns a pasted feature spec into independent coding-agent tasks with file ownership, interface contracts, a merge order and a done command for each task.
Designs a four-agent pipeline for your report type with each agent's prompt, inputs, outputs and handoff format.
Plans and executes a library or framework version migration in Cursor in small, verifiable batches, with an impact inventory, rollback points, and per-batch test gates instead of one risky bulk edit.
Designs verification checks, error handling, and a self-correction loop for an agent task before it runs, based on likely failure modes.
Runs a behavior-preserving refactor in Codex exec. It pins current behavior with characterization tests, refactors in small verified steps inside an allowed scope, and ends with an evidence-based handoff for the merge decision.
Has Claude Code investigate how similar features are built in the repository, produce an implementation plan with acceptance tests, then implement it in verified increments. Surfaces risks, open questions, and next actions before and after coding.
Interviews you about a task you do over and over, then writes a complete, save-ready SKILL.md with a trigger description, steps, output format and stop conditions.
Gives codex exec a non-interactive procedure to diagnose a failing CI output, apply a minimal fix, rerun the checks and stop after three attempts with a clear report.
Produces a PostToolUse hook entry for settings.json, the script it calls and a concrete procedure to test that the check runs after file edits.
Guides Codex through reproducing a failing CI job locally, classifying the cause as code, test, environment, or flaky, applying a minimal fix when safe, and producing an evidence-based handoff for anything it cannot resolve.
Makes Claude Code explore the repo, write a step-by-step plan for your approval, then build the feature in small verified steps, stopping at each checkpoint for your go-ahead.
Guides GitHub Copilot agent mode to inspect the repository, state assumptions, plan a minimal change, implement it, and add tests. It ends with a verification report you can use to decide whether to open a PR.
Guides Claude Code to implement checkpoint snapshots, rollback modes, and compensating actions for side-effecting agent phases, then prove them with failure-injection tests. Helps you decide the rollback strategy per phase.
A SKILL.md that makes Claude write and edit in your brand's voice: it reads a VOICE.md in the project, or builds one from samples, then checks every draft against it.
A SKILL.md that forces your coding agent to reproduce a bug with a failing test before changing code, apply the smallest fix, and prove it with the test suite.
Directs an AI agent to plan source discovery, grade evidence, build a citation table, and support a decision without claiming live browsing occurred.
Guide
Nothing that fits? Write a prompt with AI.
341 copy-ready prompts for productivity, writing, business and coding and more. Each one sets the role, the steps and the output format, and marks the parts you fill in with curly brackets.
Most are written for ChatGPT (285), Gemini (260), Claude (217) and Claude Code (42); the number is how many prompts list that tool. Each prompt page lists its tools, and most text prompts also work in any capable chat model.
Words in curly brackets, such as {audience}, are the parts you replace with your own details. Choose “Ready to use (no blanks)” in the filters to see prompts you can paste as they are.
By how often they are copied, unless you choose another order. Newest shows the latest additions first. Trending ranks prompts by copies, opens and saves over the last two weeks.
No. Every prompt is free to copy or open in your AI tool. A free account lets you save prompts to your vault.
Yes. Write with AI asks what you need, builds a short form for it, and writes the prompt as you fill it in.