Summarize
A one-sentence summary plus the takeaways worth remembering.
Summarize this in one sentence. Then list up to 5 takeaways worth remembering, only what matters most. One sentence each. No preamble, no meta-commentary.
Example prompts you can use as AI actions in Plain Markdown.
A one-sentence summary plus the takeaways worth remembering.
Summarize this in one sentence. Then list up to 5 takeaways worth remembering, only what matters most. One sentence each. No preamble, no meta-commentary.
Every concrete to-do, decision, or commitment the content suggests.
List every concrete action this content suggests — explicit recommendations, decisions, commitments, and takeaways that translate into something to do. Anchor on the primary content (post body, article, transcript). Pull from comments or replies only if they add new actionable advice; ignore conversational chatter, encouragement, and acknowledgements. Format as a checklist with "[ ]" per item, written as short imperatives. Note who an action is for if specified. If nothing actionable is present, say so in one line. No preamble, no meta-commentary.
The strongest arguments for and against, with a one-sentence verdict.
List the strongest arguments for and against the position taken in this content. Two columns: Pros and Cons. Up to 5 each, one short line per point. End with a one-sentence verdict on which side the evidence favors. No preamble, no meta-commentary.
Verify the most important factual claims against current web sources.
Search the web to verify the 3-5 most important factual claims in this content. For each claim, cross-check against current sources and rate it: Confirmed, Contested, Unverified, or False. Cite at least one source URL per claim as a markdown link in the format [domain.com](https://full-url-here). Use only URLs found in your search results — never invent URLs. If a claim cannot be verified by a found source, mark it Unverified rather than guessing. End with a one-sentence verdict on the overall reliability of this content. No preamble, no meta-commentary.
Two short replies in a forum user's voice, ready to post.
Read this discussion thread (post + top comments) and draft up to 2 reply candidates I could post under my real account. Length: 1-3 sentences, 40 words max per reply. If only one strong reply is possible, return one. Don't pad. Voice: a Reddit user typing on their phone with direct personal experience on this exact topic. Not an advisor, not an essay. Lowercase "i", contractions, and sentence fragments are fine. Open mid-thought. Never open with "Hey", "Great question", "I'm in a similar boat", "Hope this helps", or any acknowledgment of the OP or thread. End when the content ends. No summary, no call-to-action. Each reply needs one concrete detail you can stand behind: a tool name, a year, a dollar amount, a count, or a direct opinion. If you can't supply one truthfully, return fewer replies. Banned phrases (case-insensitive): moreover, furthermore, ultimately, crucial, utilize, delve, tapestry, it's worth noting, it's important to note, in today's world, plays a significant role, aims to, showcasing, aligns with, impactful, seamless, robust, cutting-edge, game-changer, incredibly, absolutely, essentially, hope this helps, happy to help. Never start a sentence with "While" or "Whether". No em-dashes, no semicolons, no three-item parallel lists. Pair one short sentence (under 10 words) with one longer one. No two consecutive sentences starting with the same word. For each reply, on the next line: "Best at: top-level" or "Best at: reply to <username>", then one line on the value it adds. No preamble, no meta-commentary.
Spot people in a thread your product genuinely helps, with peer-style reply drafts.
Step 1. Classify the thread. If this is a showcase thread, "what are you building" thread, "introduce yourself" thread, or any format where commenters are presenting their own products, projects, or work (not asking for help), output exactly: "No fit in this thread." Stop. Do not proceed to step 2. Step 2. If the thread is a real discussion where people are asking questions or describing problems, identify anyone whose problem the product below genuinely solves. <product>Replace this text with your product description: name, what it does, who it's for, what specific pain it removes. Under 80 words.</product> Match rubric (buyer-fit, not topic-fit): - Strong: user asks how to solve X, where X is the exact thing the product does, and they have not already solved it. - Possible: user describes ongoing pain matching the product's core use case, and they are not already solving it via a competitor or a homebuild. - Weak: topic overlaps but the user has already built or chosen a solution. Skip Weak entirely. A user who BUILT a tool to solve this problem is a peer, not a prospect. Skip them. Topic overlap is not buyer fit. For each surviving candidate, output: the quoted line that signals the need, the user (OP or commenter username), the match strength (Strong or Possible), and one reply draft. One reply per user maximum, even if they have multiple matching mentions; consolidate. Reply rules. Length: 1-3 sentences, 60 words max. Solve the actual problem first. Mention the product as one option among others (e.g., alongside competitors or DIY workflows). Voice: a Reddit user who actually uses this product, not a vendor. Open mid-thought. Never open with "Hey", "I had the same issue", "Hope this helps", or any acknowledgment of the OP. End when the content ends. No summary, no "this might help". Include one concrete specifier per reply: a workflow detail, a comparison, a number, or a use case. Never name the product more than once per reply. Never start any sentence with the product name. Banned phrases (case-insensitive): has been a lifesaver, is a great option, I love using, is a pain, is real, it's a different approach, gives me a lot of flexibility, I often use it to, moreover, furthermore, ultimately, crucial, utilize, delve, it's worth noting, in today's world, aims to, showcasing, aligns with, impactful, seamless, robust, cutting-edge, game-changer, incredibly, absolutely, essentially, hope this helps, happy to help. Never start a sentence with "While" or "Whether". No em-dashes, no semicolons. If nobody in this thread fits after step 2, output exactly: "No fit in this thread." Do not stretch matches. No preamble, no meta-commentary.
Same ideas, no jargon. Translates from other languages if needed.
Rewrite this in plain English for a smart adult who isn't a specialist in this topic. If the content isn't already in English, translate it in the same pass. Replace jargon, corporate-speak, and academic hedging with the clearest everyday equivalent. Keep every idea, every number, every named entity, and the original structure. Don't simplify the thinking, simplify the language. Short sentences. Active voice. Banned phrases (case-insensitive): utilize, leverage, delve, robust, seamless, holistic, synergy, paradigm, ecosystem, moreover, furthermore, ultimately, crucial, it's worth noting, in today's world, aims to, aligns with, plays a significant role. Output as Markdown. No preamble, no meta-commentary.
Every domain term defined in one short line.
Extract every domain-specific term, acronym, or technical concept a smart non-specialist might not know. For each, define it in one line of plain English, in the order it first appears in the text. Skip terms a general reader already knows. Format as a Markdown list: "**Term**: one-line definition." If nothing technical appears, say so in one line. No preamble, no meta-commentary.
The five most highlight-worthy lines, ready to share.
Pick the most highlight-worthy sentences from this text. Up to 5. Quote verbatim. No paraphrase. Each must stand alone without context: specific enough that someone reading just the line knows what it claims. Prefer sentences that are surprising, sharply worded, or carry the strongest claim. Skip filler, throat-clearing, transitions, and generic statements that could have come from any article. If the text is genuinely flat and has fewer than 5 standout lines, return fewer rather than padding. Number each quote. No preamble, no meta-commentary.
Anki-style Q&A pairs, one per concept worth remembering.
Generate flashcards from this content for spaced repetition. One card per concept, fact, or relationship worth remembering long-term. Questions test understanding, not trivia. Prefer "why does X happen" or "what does X mean" over "what year did X happen" unless the date is genuinely load-bearing. Skip anything a general reader already knows. Format each card as: Q: <question> A: <answer in one or two short sentences> Separate cards with a blank line. No numbering, no headings. If the content is too thin or anecdotal to study from, say so in one line. No preamble, no meta-commentary.
As if a friend just read it and is texting you what they thought.
Rewrite this content as if a friend just finished reading it and is texting you what they thought. 3 to 5 short messages, the way someone actually texts on their phone. Voice: casual, first-person, lowercase "i" is fine, contractions and sentence fragments are fine. The friend is sharp and well-read. They're not dumbing it down, they're just telling you what stuck. Lead with whatever's most surprising or most useful, not what came first in the article. Each message stands alone, two or three sentences max. Insert one concrete detail per message: a name, a number, or a phrase from the piece. The last message ends with their actual take, not a summary. Format as plain text, one message per paragraph. No preamble, no meta-commentary, no signatures. Banned: "in conclusion", "overall", "in summary", "to sum up", "hope this helps", "tldr", em-dashes.