Skip to content
Profile catalog

Target profiles

A profile is a recipe: the role, the structure, the safety posture, and the verification a given tool needs. Pick one in the compiler and Prompt Corner shapes your intent to fit it.

ChatGPT

chatgpt

General-purpose ChatGPT prompt. Improves clarity, structure, constraints, and output format while preserving your intent.

Best for
Everyday ChatGPT work — drafting, explaining, planning, and structured writing where you want a cleaner, better-scoped prompt.
Safety posture
Standard. No tool or file access assumed. Avoids hidden chain-of-thought and unsupported factual claims.

What it adds

  • Preserve the user's intent.
  • Improve clarity, structure, constraints, and output format.
  • Ask clarifying questions only when necessary.
  • Include citation/research instructions if current facts are involved.
  • Avoid hidden chain-of-thought requests; ask for concise reasoning summaries instead.
Rewrite with this

Claude Code

Primary
claude-code

Local coding / repo / file-agent prompt. Adds explicit working scope, audit-before-edit discipline, verification steps, and a final report.

Best for
Driving Claude Code (or a similar local coding agent) across a repository safely — audits, refactors, fixes, and reviews.
Safety posture
Conservative agent. Audit before editing, no destructive changes without approval, explicit path scope, and non-destructive verification only.

What it adds

  • Work only inside explicitly allowed paths.
  • Audit before editing.
  • Do not delete, move, rename, overwrite, or modify unrelated files unless explicitly authorized.
  • Report proposed changes before risky edits.
  • Prefer small, reviewable diffs.
  • Run non-destructive verification commands only.
  • Final report must list files inspected, files changed, commands run, failures, and remaining risks.
Rewrite with this

Codex

codex

Coding-agent prompt for OpenAI Codex-style workflows. Separates implementation, review, and verification while keeping changes minimal.

Best for
Using Codex as an implementation or second-pass review assistant on a codebase.
Safety posture
Careful agent. Tests before refactors, minimal and reviewable diffs, and no unnecessary dependencies.

What it adds

  • Treat this as a coding-agent task.
  • Identify repo structure first.
  • Prefer tests before refactors.
  • Keep changes minimal and reviewable.
  • Explain tradeoffs.
  • Do not introduce unnecessary dependencies.
  • Separate implementation, review, and verification steps.
Rewrite with this

Browser Agent

browser-agent

Browser / UI automation prompt for Claude in Chrome and similar agents. Observe-first, with confirmation before irreversible actions.

Best for
Driving a browser agent through web UIs where page state is uncertain and some actions are irreversible.
Safety posture
High caution. Observe before acting; confirm before account, payment, public-content, or irreversible changes; keep an action log.

What it adds

  • You are operating through a browser UI.
  • Do not assume page state.
  • Observe first, then act.
  • Before changing account settings, payment settings, public content, or irreversible state, report what you see and request confirmation unless explicitly preauthorized.
  • Keep a concise action log.
  • Do not handle sensitive data unnecessarily.
Rewrite with this

Research Agent

research-agent

Web research prompt with citations. Prefers primary sources, compares publication dates, and separates verified facts from assumptions.

Best for
Fact-finding and synthesis tasks where currency and source quality matter.
Safety posture
Evidence-first. Cite claims, state uncertainty, and avoid over-quoting copyrighted material.

What it adds

  • Search current sources when facts may have changed.
  • Prefer primary sources.
  • Compare publication dates.
  • Cite claims.
  • Separate verified facts from assumptions.
  • State uncertainty.
  • Do not over-quote copyrighted sources.
Rewrite with this

Image Prompt

image-prompt

Image-generation prompt. Specifies subject, composition, style, lighting, constraints, aspect ratio, and exclusions.

Best for
Producing clean, paste-ready prompts for image-generation tools.
Safety posture
Standard. Requires a reference/source when a real person or user likeness is involved; avoids unrequested text or logos.

What it adds

  • Describe subject, composition, style, lighting, constraints, aspect ratio, and exclusions.
  • Avoid extra text/logos unless explicitly requested.
  • If the image involves a real person or user likeness, require an image reference or explicit source.
  • Make the output easy to paste into image tools.
Rewrite with this

Obsidian Handoff

obsidian-handoff

Durable markdown handoff note. Captures project state, decisions, open questions, constraints, and next actions.

Best for
Writing a concise handoff note that a future AI or teammate can pick up cleanly.
Safety posture
Standard. Produces durable, structured markdown and avoids raw transcript dumps.

What it adds

  • Produce durable markdown.
  • Avoid raw transcript dumps.
  • Include project state, decisions, open questions, constraints, and next actions.
  • Keep it useful for future AI handoff.
  • Prefer concise structured notes over narrative.
Rewrite with this

App Store Workflow

app-store-workflow

App Store Connect / metadata / release prompt. Separates draft copy from confirmed production values and verifies URLs.

Best for
Preparing or reviewing App Store metadata, reviewer notes, keywords, and release checklists.
Safety posture
Verification-first. Does not invent support/privacy URLs or metadata; flags the exact fields changed or proposed.

What it adds

  • Verify URLs and app-specific metadata before treating them as confirmed.
  • Do not invent support/privacy URLs.
  • Separate draft copy from confirmed production values.
  • Include reviewer notes, metadata, keywords, privacy notes, and checklist sections when relevant.
  • Report the exact fields changed or proposed.
Rewrite with this

Coursework Helper

coursework-helper

Study and research-support prompt. Extracts, organizes, summarizes, and explains while preserving citations and academic integrity.

Best for
Accessibility-friendly study workflows: organizing, summarizing, and understanding material.
Safety posture
Integrity-first. Supports learning and accessibility; does not facilitate cheating or misrepresentation.

What it adds

  • Support accessibility and study workflows.
  • Extract, organize, summarize, and explain.
  • Do not facilitate cheating or misrepresentation.
  • Preserve citations and source boundaries.
  • Make outputs easy to read.
Rewrite with this
Found your target?Open the compiler