ChatGPT Prompts

Scan a codebase to find authentication checks that are skipped, incomplete, or conditionally disabled.

Scan code for circular imports/requires that create runtime hazards and testing friction.

Find every retry mechanism in your code and classify by strategy, timeout, and failure modes.

Diagnoses every copy, structure, and CTA failure on a landing page that is costing you signups.

Surfaces every unstated assumption baked into a technical spec before you build the wrong thing.

Turns raw customer support conversations into a ranked list of missing or broken docs.

Detects overlapping jobs, silent gaps in coverage, and jobs likely to run past their next trigger.

Finds every step in a DB migration that cannot be safely reversed and ranks the blast radius.

Surfaces every implicit trust assumption in an API or integration spec that could become a security or reliability failure.

Parses waitlist signups and any attached data to extract who wants what and why, before you build anything.

Audits a sanitized .env file to identify which variables carry secrets and how badly exposure would hurt.

Analyzes a feature backlog to surface which requests silently expand scope and erode product focus.

Stress-tests landing page copy by mapping every unanswered objection a skeptical buyer would raise before converting.

Reads a raw database schema and surfaces every structural decision that will cause data corruption, orphans, or silent failures at scale.

Converts a dump of support tickets into a ranked map of missing features, broken promises, and UX debt.

Expose what a job posting actually reveals about team dysfunction, unclear ownership, and unrealistic expectations.

Paste a live API response and a schema or docs excerpt to find every field that lies, drifts, or goes missing.

Analyze what a founder's internal update reveals about team stress, misalignment, and unspoken pressure beneath the optimism.

Reads a spec and reorders the build sequence by where unknown risks and integration traps are hiding.

Takes raw churn data or cancel reasons and builds a ranked hypothesis map with investigation steps.

Separates the metrics you track out of habit from the ones that actually predict outcomes.

Finds the exact moments where your landing page breaks the visitor's mental model and loses the sale.

Converts an ambiguous business goal into a structured experiment with a clear pass/fail condition.

Finds every question in your support history that shouldn't exist if your docs were good enough.

Give the model a function signature and it generates the invariants, edge cases, and properties worth testing.

Paste competitor onboarding emails and get a full funnel map with psychological triggers and gaps.

Paste a batch of competitor job listings and extract strategic signals about their roadmap and weak spots.

Critiques a dashboard image by asking what decisions each chart actually enables and which ones are decorative.

Converts raw customer interview transcripts into a structured JSON database of pains, workarounds, and willingness-to-pay signals.

Analyzes your onboarding email sequence to find where users disengage and why your copy isn't pulling them through.

Drop a Stripe export into Advanced Data Analysis and get the cohort retention story your dashboard is hiding.

Have GPT-4o browse a competitor's public changelog and reverse-engineer the strategy behind their ship cadence.

Feed raw production logs to GPT-4o and get a ranked list of failures that aren't paging anyone yet.

Turns messy Slack threads and timeline scraps into a structured incident report that finds systemic causes.

Feed homepage screenshots of your competitors and get a structured map of where positioning is crowded vs open.

Strict schema that forces GPT-4o to convert your week's notes into structured decisions with reversibility scores.

Uses image understanding to diagnose every decision friction on a pricing page, ranked by likely revenue impact.

A system prompt that turns ChatGPT into a brutal line editor for product copy, with diffs and reasoning, not rewrites.

Turns a raw analytics export into a narrated funnel breakdown that flags noise, segments, and false signals.

A strict JSON output contract for code review agents that separates findings, evidence, and confidence so a downstream pipeline can act on them.

Analyzes a UI screenshot for cognitive load, hierarchy failures, and the specific decisions a user is forced to make.

Turns raw product analytics events into a ranked diagnosis of where and why new users abandon onboarding.

Examines a dashboard image and flags which metrics are decision-grade vs which are vanity or likely misleading.

Crawls competitor changelogs and release notes to infer strategic direction, dying features, and timing patterns.

Analyzes a Stripe CSV export to separate durable revenue from fragile revenue with concentration risk flags.

A system prompt that forces ChatGPT to extract, validate, and self-grade structured data with confidence scores.

Paste raw server logs and get ranked failure hypotheses with evidence citations and next diagnostic steps.

Point ChatGPT at a competitor's careers page and get a structured read on their roadmap and priorities.

Send ChatGPT to a niche subreddit and return a structured demand signal report ranking real problems by buying intent.

Upload business bank statements and get a categorized burn analysis plus a 6-month runway forecast with scenario branches.

Upload a UI screenshot and get a ruthless IA audit grounded in visual hierarchy and cognitive load principles.

Browses a competitor's changelog or release notes and extracts strategic signals about their roadmap and positioning.

Analyzes a UI screenshot for specific conversion blockers using behavioral and visual hierarchy principles.

Uses Code Interpreter to mine refund and dispute data for fraud signatures and high-risk customer patterns.

Force ChatGPT to return strictly valid JSON matching your schema, with self-validation and a fail-safe.

Build a persona grounded in actual evidence, with confidence levels and explicit gaps marked.

Run a real cohort retention and revenue analysis on your Stripe CSV with Code Interpreter, no SQL needed.

Browses competitor pricing pages and returns normalized structured JSON ready for a comparison spreadsheet.

Forces ChatGPT's Advanced Data Analysis to find what is wrong with your data, not just describe it.

Adversarially probes your system prompt for ambiguity, contradictions, and edge cases before users find them.

Upload any CSV and get a full data quality report with column types, nulls, outliers, and cleanup code.

Rewrite a pricing page to remove friction and surface the right tier for each visitor.

A spec for a high-converting SaaS landing page — section by section, with the questions each must answer.

A launch announcement that sounds like a founder, not a press release.

Generate a tailored launch checklist for a specific product/feature, not a generic template.

A staged rollout plan that protects production while shipping new behavior.

Pressure-test a proposed pricing model before you ship it.

Get a tight, founder-style positioning doc instead of a watered-down marketing brief.

Generate a useful postmortem from logs, timeline, and chat history.

Cold email/DM that does not sound AI-generated. Built for founders pitching real work.

Get a real, opinionated tradeoff analysis instead of "it depends". Useful for technical decisions.

Walk through a bug with ChatGPT in a way that surfaces what you missed instead of just generating a fix.

A structured prompt that forces ChatGPT to design a system the way a staff engineer would, not a tutorial.