120 “Secret Codes” for ChatGPT That Give You Better Answers
Paste one short cue before your prompt to get clearer writing, sharper research, faster plans, and better work.
I went looking for the giant list behind Sabrina Ramonov’s viral Reel.
Her public Substack post has 19 fun ChatGPT codes. The resource linked under it is much bigger. I found 935 public prompt and resource files in the linked GitHub repository when I checked it on July 14, 2026.
I rebuilt the idea from scratch for Liz on the Web readers. Below are 120 short prompt cues for clearer answers, sharper research, faster plans, better writing, and work you can use.
These are user-made prompt cues. They are not built-in ChatGPT commands or hidden OpenAI features. The code is a label that helps you remember an instruction.
OpenAI’s own prompt guide says clear, specific instructions with enough context tend to work best. That means a code by itself may not do much in a fresh chat. The first time you use one, paste the code and its one-line meaning before your request.
HOW TO USE THEM
Try this shape:
GOALLOCK: Restate my goal, key facts, and limits before answering. Then help me plan a 30-day launch for my new offer. I have two hours a day and a $500 budget.
Save the codes you use most in a note, project, or custom instruction. Add your real goal, audience, facts, limits, and wanted format after the code.
CLEAR ANSWERS
1. GOALLOCK: Restates your goal, key facts, and limits before answering.
2. PLAINWORDS: Rewrites the answer with short words and simple sentences.
3. ANSWERFIRST: Puts the direct answer before the reasons and steps.
4. SCOPECHECK: Flags any part of the request that is unclear or too broad.
5. ASSUMPTIONS: Lists each assumption so you can correct it.
6. EXAMPLEPAIR: Gives one good example and one bad example.
7. STEPSTACK: Turns the answer into ordered steps with one action per step.
8. SHORTCUTVIEW: Gives the fastest workable path before a more complete path.
9. TERMSCAN: Finds jargon and explains each term in plain words.
10. FORMATLOCK: Returns the answer in the exact format you name.
RESEARCH AND FACT CHECKS
11. SOURCEFIRST: Separates checked facts, likely claims, and open questions.
12. DATESTAMP: Adds the date when each time-sensitive fact was checked.
13. CLAIMSCAN: Marks claims that need proof and names the best source type.
14. SOURCEGRADE: Ranks sources by trust, freshness, and direct evidence.
15. GAPFINDER: Shows what evidence is missing before reaching a conclusion.
16. LINKMAP: Pairs each key fact with the source that supports it.
17. UPDATECHECK: Finds advice that may be old or changed.
18. QUOTECHECK: Checks a quote against the source you provide.
19. NUMBERTAG: Adds a source, unit, and context to every number.
20. COUNTEREVIDENCE: Searches for facts that could weaken the current view.
IDEAS AND PLANNING
21. ANGLEGRID: Creates ideas based on audience pains, goals, and content types.
22. IDEALADDER: Builds small, medium, and bold versions of one idea.
23. OPTIONTRIO: Gives a safe option, a fast option, and a bold option.
24. NEXTMOVE: Picks the best next action based on impact, time, and risk.
25. PLAN30: Builds a 30-day plan with weekly goals and daily actions.
26. BACKPLAN: Starts with the end goal and works backward to today.
27. BLOCKMAP: Lists blockers, needed work, owners, and due dates.
28. TESTFIRST: Suggests the smallest useful test before a larger commitment.
29. PRIOSTACK: Ranks tasks by impact, effort, and urgency.
30. STOPLIST: Finds low-value work that can be paused or removed.
WRITING AND EDITING
31. TONEFIT: Rewrites the text for the audience and tone you name.
32. CUT20: Cuts the draft by about 20 percent without losing its meaning.
33. STRONGVERBS: Replaces weak phrases with clear, active verbs.
34. OPENHARD: Writes a stronger first line based on the main tension.
35. FLOWFIX: Reorders the draft so each idea leads into the next.
36. REPEATSCAN: Removes repeated points while keeping needed context.
37. PROOFPASS: Fixes spelling, grammar, and punctuation without changing the voice.
38. VOICEKEEP: Improves clarity while keeping the writer’s natural style.
39. CTAFIT: Writes one clear call to action for the next best step.
40. TITLEBANK: Creates 10 clear titles from one core promise.
CONTENT AND SOCIAL MEDIA
41. HOOKGRID: Creates hooks based on curiosity, stakes, proof, and speed.
42. REELBEATS: Turns an idea into a hook, three beats, and a CTA.
43. CAROUSELMAP: Splits a lesson into slides with one main idea per slide.
44. CAPTIONTIGHT: Writes a short caption that adds context to the post.
45. SERIESMAKER: Turns one topic into a repeatable content series.
46. REPURPOSE5: Adapts one source into five different content formats.
47. COMMENTASK: Writes a natural question that can lead to useful replies.
48. CONTENTGAPS: Finds unanswered questions in comments, reviews, or FAQs.
49. PROOFPOINTS: Lists the examples and facts needed to support a post.
50. PLATFORMFIT: Adjusts the length and structure for the chosen platform.
BUSINESS AND OFFERS
51. PAINMAP: Maps customer pains to goals, barriers, and possible help.
52. OFFERCORE: States who the offer helps, what it does, and how it works.
53. BENEFITSHIFT: Turns a feature list into clear buyer outcomes.
54. OBJECTION10: Lists 10 buyer concerns and a useful response to each.
55. PRICESTORY: Explains price through scope, saved time, and expected benefit.
56. NICHECHECK: Tests a niche for a clear buyer, urgent need, and reachable market.
57. OFFERTRIO: Creates starter, main, and premium package ideas.
58. PROMISECHECK: Flags claims that are vague, risky, or unsupported.
59. DELIVERYMAP: Turns an offer into repeatable steps and needed assets.
60. RETENTIONPLAN: Finds ways to improve the customer’s long-term experience.
SALES AND CUSTOMERS
61. LEADFIT: Scores a lead against the rules you provide.
62. DMSTART: Drafts a short opener based on real details about the person.
63. FOLLOWUP3: Writes three follow-ups with a different goal for each.
64. CALLPREP: Builds a research list and useful questions for a sales call.
65. NEEDSNOTES: Turns call notes into pains, goals, concerns, and next steps.
66. PROPOSALONE: Writes a one-page proposal with scope and price placeholders.
67. REPLYCALM: Drafts a calm response to an upset customer without placing blame.
68. FAQBUILD: Turns real customer questions into a clear FAQ.
69. WINLOSS: Compares won and lost sales to find patterns.
70. REFERRALASK: Writes a brief referral request with a clear reason.
WORK AND DECISIONS
71. DAYFOCUS: Picks the three most useful outcomes for the day.
72. TIMEMATCH: Fits tasks into the time and energy you have.
73. DECISIONCARD: Shows the choices, rules, risks, and best option.
74. RISKLOG: Lists risks with their chance, impact, and next action.
75. DELEGATEIT: Finds work to hand off and the details the new owner needs.
76. CHECKPOINTS: Adds review points so problems can be found early.
77. TWOMINUTE: Pulls out tasks that can be finished in two minutes.
78. BATCHWORK: Groups similar tasks to reduce task switching.
79. WEEKRESET: Reviews wins, misses, lessons, and next priorities.
80. DONECHECK: Defines what finished work must include before work begins.
LEARNING AND COACHING
81. TEACHBACK: Asks you to explain the idea and then fills the gaps.
82. SKILLPATH: Builds a learning path with lessons and small practice projects.
83. QUIZLOOP: Asks one question at a time and changes the level as you improve.
84. ERRORCOACH: Explains why an answer failed and how to avoid the same mistake.
85. MEMORYCARD: Turns study material into short recall questions and answers.
86. PRACTICESET: Creates practice tasks that move from easy to hard.
87. ANALOGYFIT: Explains an idea with a familiar example and shows its limits.
88. STUDY20: Builds a 20-minute study session with recall and review.
89. RUBRICMAKER: Creates a clear scoring guide for a skill or assignment.
90. FEEDBACKPLAN: Turns feedback into small fixes in priority order.
MEETINGS AND TEAMS
91. AGENDAFAST: Builds a timed agenda from the meeting goal and guest list.
92. PREMORTEM: Lists ways a plan could fail and how to reduce each risk.
93. ROLECLEAR: Names the owner, decision maker, helpers, and deadline.
94. NOTE2TASK: Turns meeting notes into decisions, tasks, owners, and dates.
95. DECISIONLOG: Records a choice, its reason, its evidence, and a review date.
96. ASYNCFIRST: Turns a meeting topic into a written update and key questions.
97. CONFLICTMAP: Separates facts, needs, guesses, and possible fixes.
98. STATUSONE: Shrinks a project update into done, next, and blocked.
99. HANDOFFKIT: Creates a handoff with context, files, risks, and the first step.
100. RETRO3: Runs a simple keep, stop, and try review.
DATA AND DOCUMENTS
101. TABLECLEAN: Fixes messy headers, values, dates, and number formats.
102. DATAGAPS: Finds missing, repeated, or invalid values.
103. TRENDREAD: Explains changes and possible causes without treating guesses as facts.
104. CHARTPICK: Picks a chart based on the question and data shape.
105. FORMULACHECK: Reviews spreadsheet formulas and flags likely errors.
106. DOCMAP: Outlines a long document with one key point per section.
107. COMPAREDOCS: Compares two documents for changes, conflicts, and missing parts.
108. POLICYPLAIN: Rewrites a policy as simple rules and steps.
109. EXTRACTFIELDS: Pulls named details from text into a structured table.
110. JSONSHAPE: Turns supplied information into valid JSON with your field names.
TECH AND BUILDING
111. BUGTRIAGE: Turns a bug report into repeat steps, likely causes, and tests.
112. ROOTPATH: Traces a problem to likely root causes before suggesting a fix.
113. TESTMAP: Lists the highest-risk cases and the right tests for them.
114. SPEC2TASKS: Breaks a feature plan into small tasks with completion checks.
115. CODEPLAIN: Explains code through its inputs, outputs, steps, and risks.
116. DIFFCHECK: Reviews a code change for bugs, broken behavior, and missing tests.
117. APISCHEMA: Drafts request and response shapes from your requirements.
118. EDGECASES: Finds empty, invalid, repeated, timeout, and permission cases.
119. RELEASEGATE: Builds a release checklist for tests, settings, rollback, and monitoring.
120. DOCSTART: Creates a short README plan with setup, use, limits, and fixes.
THE FAST WAY TO START
Pick five codes that match work you already do. Save the full line, not only the label. Add your real context after it. Then keep talking to ChatGPT until the answer fits the job.
My five beginner picks are GOALLOCK, ANSWERFIRST, UPDATECHECK, NEXTMOVE, and DONECHECK.
Credit where it is due: Sabrina Ramonov’s Reel sent me looking for the larger resource. Her public post shares 19 of her own codes and links to a much larger public prompt library. I used that idea as the spark, then wrote this 120-code list from scratch.
SOURCES
Sabrina’s Reel:
Sabrina’s 19-code post: https://www.sabrina.dev/p/secret-codes-for-chatgpt
Sabrina’s public prompt library: https://prompts.sabrina.dev/
Linked GitHub repository: https://github.com/SabrinaRamonov/prompts
OpenAI’s ChatGPT prompting guide: https://help.openai.com/en/articles/10032626-how-do-i-prompt-chatgpt-effectively
KEEP GOING
My plain guide to GPT-5.6 Sol, Terra, and Luna: https://www.lizontheweb.com/p/gpt-56-is-here-sol-vs-terra-vs-luna
My five-model comparison: https://www.lizontheweb.com/p/5-ai-models-compared-after-todays
My six-skill AI content system: https://www.lizontheweb.com/p/6-ai-content-skills-that-can-plan-d8f
Which code should I turn into a full copy-and-paste prompt next? Tell me in the comments.