> For the complete documentation index, see [llms.txt](https://thecontentforge.gitbook.io/thecontentforge-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://thecontentforge.gitbook.io/thecontentforge-docs/feature-glossary/glossary/content-forge.md). # Content Forge *Last updated: 2026-08-07.* Content Forge is where writing happens. You pick a generation mode, give it something to work from, and get back finished posts written in your brand voice and shaped for each platform you publish to. Which modes you see depends on the kind of workspace you are in, so a real estate team and a crypto project do not wade through each other's tools. ## Core modes (available to every workspace) The general purpose writing modes. Every workspace gets these, with one exception noted under the meme and culture group below. * **Rewrite Draft** - Paste a rough draft and get it rewritten in your brand voice, with a separate version tuned for X, Instagram, and Facebook. This is the mode for when you know what you want to say but not how to say it well on three networks at once. * **Breaking News** - Paste an article, a link, or raw news text and get posts built around it that are ready to publish. A Breaking switch tells it to write with the urgency of a story that just landed. It finds the angle that matters to your audience rather than just summarising the story. * **Image Post** - Upload a chart, an infographic, or a screenshot and get captions written for each platform based on what is actually in the picture. You can add a line of context to steer how it reads the image. * **Post Helper** - Give it a topic, a data point, or an excerpt and it builds an original on-brand post around it. This is the general starting point when you have an idea but no draft. * **Thread** - Turns one idea into a multi-part X thread, typically three to ten parts, each part sized to the X character target you set. Instagram and Facebook get normal single posts covering the same ground. * **Comment King** - Generates three to four X reply options for someone else's post, each taking a different angle such as a thought-provoking take, a genuinely useful addition, and a respectful counterpoint. It also names its best pick and comments on timing, so you can grow through replies instead of only through your own posts. * **What's Working** - A read-only view of up to six performance pattern cards for your recent posts, broken down by platform, with median engagement rate, median impressions, the hook types that landed, and a warning when a metric has shifted. Nothing is generated here, it is the evidence behind everything else. The full pattern engine it reads from is on the Content Optimization page. * **Brand Voice** - Shows the active brand voice profile the AI applies to every generation in this workspace: tone rules, things to do more of, things to avoid, signature moves, and preferred and avoided words. Open it when you want to understand why the output sounds the way it does. It is a viewer here; the voice itself is built and edited on the Content Optimization page. ## Newsletter A complete long-form email builder that sits inside Content Forge. You choose what the issue is made from, it drafts the whole thing, and you edit, illustrate, and export it without leaving the page. * **Sources** - Choose exactly what the issue is built from, up to twenty-five items. You can pull recent stories from your tracked feeds, add article links, add X posts, or paste in your own notes, and you can also pick from links and posts already collected in your workspace. Drag them into the order you want, because that order becomes the order of the sections in the finished email. * **Smart pick** - Fills the source list for you in one click with a strong, balanced selection of recent feed items, so the issue opens with fresh, newsworthy material and no single site dominates it. * **Deep pick with AI** - A wider, slower curation pass that looks across everything the workspace tracks, including feeds, competitor posts, your own brand accounts, and your own published posts, then proposes a balanced set with a short reason attached to each choice. Use it when you want the issue curated rather than merely filled. * **Newsletter setup** - Set the issue type (weekly digest, project update, market roundup, research brief, launch recap, or custom), how long each section runs, the tone, and who it is written for. Separate switches control whether images, citations, a sources block at the bottom, and email-ready HTML are included. You get one section per source you selected, so section count follows the source list rather than a preset. * **Drafting a full issue** - Writing a full issue takes a few minutes, and a live progress panel shows how far along it is. You can cancel it or start it again without rebuilding the source list, and leaving the page or refreshing does not lose the work. * **Newsletter editor** - Everything the draft produced is editable. That covers a choice of subject lines, the preview line, brand name, logo and colours, the hero, the intro, each section's heading, summary, body, bullet points, byline and date, plus short round-up items. Sections and call-to-action blocks sit in one shared order, so anything can be moved up, moved down, or removed. * **Section and hero images** - Every section can carry an image lifted from its source article, chosen from your asset library, or generated on the spot, and a generated image can then be revised by describing the change in plain words. The hero image works the same way and also accepts a direct upload. * **Preview and export** - See the issue rendered as an email, as plain text, or as raw HTML, then copy either version or download it as a file. This is what you paste into whichever tool actually sends your mail. * **Saved newsletters** - Save an issue and it is kept so you can reopen and carry on later. Saving again creates a new version rather than overwriting the old one, and past issues are listed with their images so you can find one at a glance. ## Web3 project modes These two appear for workspaces set up as a Web3 project. * **Summarize the project** - Builds a promotional X thread about your project entirely from project knowledge you have already approved, so every claim traces back to something you signed off on. If the project profile is too thin to do that safely, it stops and tells you exactly what to fill in first instead of inventing detail. Instagram and Facebook versions adapt automatically. * **Changelog Update** - Turns release notes into announcements. Point it at up to five changelog pages, paste the notes, or both, and it pulls the release into a structured list you can review and correct before a word is written. You then pick the target platforms from X, LinkedIn, Facebook, Telegram, Threads, Instagram, and Discord, set the audience and the technical depth, and it writes the announcement for each one. ## Meme and culture modes These appear for meme and culture coin workspaces. Those workspaces get this set in place of the standard writing modes, keeping only the Brand Voice viewer from the core group. Everything here is grounded in the workspace's own meme personality, and about half the modes return a toned-down safer version next to the main output so you always have a fallback. * **Meme Post** - Short, native X posts in your coin's voice. You can optionally set the angle from Degen, Absurd, Bullish, Cult-like, Sarcastic, Community rally, Mascot POV, Chart reaction, Reply bait, and Quote-tweet reaction. You get the main post plus alternate hooks, a Telegram version, an image idea, and a safer version. * **Community Rally** - Wakes up Telegram and holders, whether that means pushing a coordinated moment, celebrating a milestone, or turning a red candle into a joke. Returns a main X post, a Telegram version, a set of short replies, and a safer version. * **Community Push Kit** - Bulk engagement material for your community to use. You can pick the output type from ten short replies, five quote-tweet angles, three community push prompts, five caption ideas, or five lines designed to sit under a large account. It is written to stay clear of spam, harassment, and brigading. * **Lore Thread** - Turns a thin concept into lore the community can actually use. Returns a thread, a pinned post, a Telegram welcome message, and an image idea. * **Trending Reaction** - For reacting fast when something goes viral or the market moves. Returns a reaction post, a quote tweet, a pack of short replies, an image angle, and a safer version. * **Milestone Hype** - Announcements for holder counts, volume, market cap, new listings, or influencer mentions. Returns an X post, a Telegram post, short replies, an image idea, and a more restrained backup version. * **Chart Reaction** - Turns a price move into content. Returns an X post, a quote tweet, an image prompt, and a safer version. * **Telegram Announce** - Writes a Telegram-native announcement plus a mirrored version for X, so both channels say the same thing in the right register for each. * **Pinned Post** - Writes a pinned post for X, a matching pinned message for Telegram, and an image prompt for the cover art. * **Image Prompt Pack** - Produces five to eight image generation prompts covering memes, banners, profile pictures, sticker ideas, and comic panels, giving a designer or an image tool a full batch to work from. * **CTO Relaunch** - Messaging for a community takeover, with extra care around what can and cannot be said about the previous team. Returns an announcement, a trust-rebuilding post, a rally thread, a fresh-start pinned post, a Telegram announcement, an answer for the obvious questions, and a safer version. * **Influencer Replies** - Reply variants written to sit under large accounts, plus quote-tweet angles, both aimed at borrowing reach from bigger conversations. * **Meme Campaign** - Plans five to seven sequenced posts as a single campaign arc, each with its own image prompt, so a stretch of posting is decided in one sitting. * **Weekly Recap** - Pulls the week together into a recap thread, a Telegram recap, and a short summary post for X. ## Real estate modes These appear for real estate workspaces. Every one of them is generated from a property you select, so the facts and photos come from your own verified listing record rather than from the model's imagination, and the writing is steered away from fair housing problems and unsupported claims. * **Listing Description** - Writes the listing copy from the property record as a headline, a full description of two to four paragraphs, and a short teaser. It is built to be safe to submit to a listing service. * **Coming Soon** - A teaser that builds interest before the listing goes active. * **Just Listed** - A high-energy announcement that the property is on the market, with a clear next step for buyers. * **Feature Post** - Spotlights a single room or feature and anchors the writing to the photos of the room you choose. * **Open House** - Promotes an open house. You have to supply the date and time yourself, it will never invent one. * **Price Improvement** - Reframes a price change as a fresh opportunity so buyers who passed the first time take another look. * **Just Sold** - Celebrates the result and turns it into social proof with a light prospecting ask attached. * **Neighborhood Spotlight** - Covers amenities, lifestyle, and access around the property, kept factual and free of steering language or unverified school claims. * **Tour Video Script** - A scene-by-scene walkthrough script for a property tour, with on-screen text and voiceover lines. * **Market Update Script** - A short script for an agent's market update video. * **Rental Listing** - A rental description covering rent, availability, and amenities. * **Rental Social Post** - Announces that a unit is available and invites people to book a viewing. * **Amenity Post** - Focuses on one amenity, anchored to the space you pick. * **Rental Video Script** - A walkthrough script for a rental unit. * **Showing Invite** - A concise invitation to come and see the place. * **Rental Follow-up** - A follow-up message after a rental enquiry or a viewing. * **Application Next Steps** - Spells out what happens next for a renter who wants to apply. * **Tenant Welcome** - A warm move-in message for a new tenant. * **Lease Renewal** - Renewal outreach to a tenant whose lease is coming up. * **Owner Update** - A status and activity update written for the property owner. * **Maintenance Update** - A maintenance update for the owner, written to avoid admitting liability. * **Portfolio Update** - A roll-up update covering several properties at once. * **Investor Update** - A performance update on one property, written for an investor. * **Vacancy Post** - Fills a vacancy with urgency that stays believable rather than pushy. * **Outreach** - One mode covering direct, situational messages. You choose the audience from eighteen options spanning seller prospects near a recent sale, buyer and open house follow-ups, investors, past clients, expired and for-sale-by-owner contacts where that is legal, and the full renter and owner lifecycle. Then you choose the channel: email, text message, direct message, call script, voicemail, letter, or a note for your records. * **Only the modes that fit the property** - The picker greys out any property whose stage or purpose does not match the mode, and tells you why, so you cannot write a just-sold post for a listing that is still active or a lease renewal for a home that is for sale. Archived properties cannot produce new content at all. * **Finding a mode fast** - Real estate workspaces get a grouped picker (listing marketing, rental and owner, video scripts, outreach, general) with a search box and a shortlist of the content types you used most recently. Twenty-five modes is a lot to scroll, so this is how you get to the one you want. * **Compliance check on results** - Real estate output comes back with a verdict of Approved, Needs verification, or Do not publish, and where something is flagged it shows the exact phrase and suggests safer wording. It is a second pair of eyes before anything goes out. ## Student modes These appear for student workspaces, and which ones show up depends on the lanes the student chose during setup. They are labelled Beta in the product (beta / in progress). Generations spend credits from a monthly allowance, and the cost is shown next to the generate button before you run anything, from one credit for a single post up to ten for a full kit. Brand Voice and What's Working are free to open, and the newsletter builder is not metered. * **Student Creator** - For building a personal following. Covers single posts with platform-specific variants, caption and hashtag packs, seven-day and thirty-day content plans, bios tuned for each platform, and a full personal brand kit with positioning, a hook bank, and content pillars. * **Career & Internships** - For landing roles and looking credible while doing it. Covers LinkedIn posts, new-role announcements across several platforms, networking and recruiter messages, career fair follow-ups, and a bio set covering a LinkedIn about section, a resume summary, and a short introduction. * **Clubs & Campus Events** - For running a club, a chapter, or an event. Covers event promotion packs, recruitment material including a tabling pitch and outreach templates, member spotlights, fundraiser posts, event recaps, and a full campaign kit spanning tease, announcement, day-of, and recap. * **Student Entrepreneur** - For launching something. Covers founder story posts, launch day announcements, a customer persona with sample messaging, a seven-day launch plan, pitch copy in three lengths, and a full brand launch kit. * **Athlete / NIL** - For athletes managing their own name and image. Covers game day and training posts, athlete bios, sponsor-friendly posts with a more conservative alternative alongside, deal announcements with the required disclosure wording, and a positioning kit for approaching partners. * **Video & Podcast** - Choosing this lane does not add modes to Content Forge. Its episode titles, clip hooks, descriptions, and repurposing tools live in the video product instead, where they can work from the actual recording rather than a description of it. * **Guardrails on every student result** - Every generation carries a reminder that the draft is AI-assisted and that the student is responsible for their school's name-and-image policy, their code of conduct, and disclosing paid partnerships. When the draft had to be softened, or a paid-partnership disclosure had to be added, you get a note saying so rather than a silent change to the text. ## Working in Content Forge The parts of the writing surface that apply no matter which mode you are in. * **Shape it before you generate** - Above the generate button you set the target platform, the tone, the goal of the post, and a character target for the X version, either one of the presets or your own number. Web3 workspaces add a content type selector and extra tones, and student workspaces add their own tones. * **History** - Keeps your recent generations, currently the last fifteen, and lets you reopen any of them complete with their images, or delete the ones you do not want kept. Each entry shows the mode, the original input, a preview of the output, and, where one was given, the brand voice score out of a hundred. * **Generations survive you leaving** - Closing the tab, refreshing, or switching to another mode does not lose a generation in progress. Results come back into the slot for the mode and platform you ran them from, and a small panel shows anything still running. * **Load a URL as a source** - Paste any article, blog post, or X link and it reads the page for you, then drops a clean summary with the title, author, key claims and figures into the writing box. For an X link you get the post itself rather than the page around it, and pasting the same link again brings the summary straight back. * **Drop in a screenshot** - Drop, paste, or browse to a screenshot and it reads the text out of the image and turns it into a source you can write from, and tells you how confident it was. Useful for a chart, a group chat message, or anything you can only get hold of as a picture. * **Send in a source from elsewhere in the product** - Knowledge base pages, trend scans, pattern findings, token research, asset library items, and a competitor post you want to answer can all be pushed straight into the writing box from where you were reading them, sometimes carrying a suggested mode and tone with them. If the source is longer than the box holds you are told exactly how much made it in, and the whole document is still taken into account while the post is written. * **Quote someone else's post** - Tick the box and paste the exact words of the post you are responding to, and they will be quoted verbatim rather than paraphrased. This removes the risk of putting words in someone else's mouth. * **Generate headline images** - A switch on the text-driven modes that produces a headline image alongside the written post, wide for X and Facebook and square for Instagram, matched to what the post actually says. You can revise any image afterwards by describing the change, download it, and it is filed into your asset library automatically. * **Revise any line without regenerating** - Every generated block, including each meme output card, has a revise button: describe the change in plain words and only that piece is rewritten. The rest of the result stays exactly as it was. * **Fast or Best** - Choose between a quick draft and a slower option that delivers a noticeably more polished result. Fast is for volume, Best is for anything that matters. The slower option leaves behind an expandable summary of what it changed and the brand voice score it settled on. * **Algorithm signals** - When the workspace has platform algorithm sources configured, results show how many signals and sources were applied, a per-platform score, and the exact phrases in the copy that a signal shaped, highlighted in the text. Hover a highlight to see which signal it came from. * **X engagement suggestions** - Alongside an X post you may get suggested cashtags, accounts worth tagging, accounts worth quoting to amplify, and posts worth replying to for visibility. Each suggestion comes with a short note on why. * **Live Price Snapshot** - When your input mentions tokens, real prices are pulled in and shown next to the result, along with anything that could not be found. It keeps numbers in the post honest and shows you what the writing was actually based on. * **Publish or schedule from the result** - Once you have output you are happy with, you can send it to a connected account, schedule it for a specific date and time, or just mark it as posted manually. Multi-part threads are queued as a linked chain on X and folded into one post on Instagram and Facebook, so you never have to copy a result somewhere else to get it out. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://thecontentforge.gitbook.io/thecontentforge-docs/feature-glossary/glossary/content-forge.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.