> 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-guides/content-studio.md). # Content Forge Content Forge is where you generate draft content across your platforms. It combines AI generation with your organisation's brand voice so drafts come out close to publish-ready. > **Before you generate your first post:** set up your [Brand Voice](/thecontentforge-docs/feature-guides/brand-voice.md) first. Content Forge writes every draft in your Brand Voice, so creating it is a required setup step - without it, drafts fall back to a generic tone. The recommended order is: finish the Setup Wizard, create your Brand Voice, add Knowledge Base sources, connect your social accounts, then generate. ### Generation modes Content Forge has nine standard modes plus one Web3-only mode, and a separate family of **meme / culture-coin modes** that appears for meme orgs (see below). Each is purpose-built for a specific input type or workflow: | Mode | What it does | | ------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Rewrite Draft** | Paste a draft post and get on-brand rewrites for X, Instagram, and Facebook | | **Breaking News** | Paste an article, URL, or news text to generate ready-to-post content | | **Image Post** | Upload a chart or infographic; the AI writes captions for each platform | | **Post Helper** | Share a topic, data point, document summary, or article; the AI crafts an original on-brand post | | **Thread** | Create a 3-10 part X thread for deeper analysis; Instagram and Facebook receive normal platform posts | | **Comment King** | Generate 3-4 high-value X reply options to grow engagement on top posts | | **What's Working** | View current performance patterns and top hooks from your post history | | **Brand Voice** | View your active Brand Voice DNA | | **Newsletter** | Build a long-form, email-ready newsletter from RSS, articles, X posts, and notes - one section per source. See [Newsletter](/thecontentforge-docs/feature-guides/newsletter.md) | | **Summarize the project** | *(Web3 Project orgs only)* Promo-grade X thread assembled from your approved Project Knowledge - profile, differentiators, approved claims, team, roadmap, FAQs, and other approved supporting context. IG/FB adaptations are generated alongside the X thread. | ### Meme & culture-coin modes Orgs configured as a **meme / culture coin** account get an additional family of modes built for community-driven, fast-moving meme accounts. They're driven by your **Meme DNA** and return *structured, multi-part* outputs - typically a main X post, a Telegram version, an image prompt, and often a **safer version** you can fall back to. | Mode | What it does | | ---------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | **Meme Post** | Short, native X posts from your Meme DNA. Pick a sub-mode - Degen, Absurd, Bullish, Cult-like, Sarcastic, Mascot POV, Chart reaction, Reply bait, Quote-tweet reaction - for the angle | | **Community Rally** | Wake up Telegram and holders - push community moments, celebrate milestones, turn red candles into comedy | | **Community Push Kit** | Replies, quote-tweet angles, and community-push prompts. Sub-modes for short replies, quote-tweet angles, push prompts, captions, and reply-under-big-account. No spam, harassment, or brigading | | **Lore Thread** | Build usable lore from a thin concept - outputs a thread + pinned post + Telegram welcome | | **Trending Reaction** | React fast to viral moments and market events - post + quote-tweet + reply pack + image angle | | **Milestone Hype** | Celebrate holders, volume, market cap, listings, or mentions - with a safer-version backup | | **Chart Reaction** | Turn a chart move into meme content - X post + quote-tweet + image prompt + safer version | | **Telegram Announce** | A Telegram-native announcement with an X mirror version | | **Pinned Post** | X pinned post + Telegram pinned version + a cover image prompt | | **Image Prompt Pack** | 5–8 image-generation prompts for memes, banners, PFPs, sticker ideas, comic panels | | **CTO Relaunch** | Community-takeover messaging with extra trust guardrails - transparent-rebuild language, avoids defamation | | **Influencer Replies** | Reply-under-big-account variants + quote-tweet angles for visibility | | **Meme Campaign** | 5–7 sequenced posts with a campaign arc + per-post image prompts | | **Weekly Recap** | A weekly community recap thread + Telegram recap + X summary | These modes pull from the same Meme DNA used across [Meme Intelligence](/thecontentforge-docs/feature-guides/meme-intelligence.md). The **safer version** and the guardrail language on Community Push Kit and CTO Relaunch are deliberate - they keep hype on the right side of platform rules and avoid claims you can't stand behind. ### Source inputs You can start from a draft, a topic, a URL, a screenshot, an uploaded image, or a document sent from the [Asset Library](/thecontentforge-docs/feature-guides/asset-library.md). * **URL loader** - paste an article, blog post, or X post URL to load it as source material before generating * **Screenshot loader** - drop, paste, or browse for an image; the extractor pulls out text, entities, claims, and context for Post Helper * **Referenced post** - when you are making a post around someone else's post, mark that workflow and paste the source handle or text so the quote context stays locked * **Document source** - click **Use as source in Content Forge** from a PDF in Asset Library to prefill Post Helper with summary, claims, metrics, and angles * **Intelligence handoffs** - the Intelligence pages can push context straight into Content Forge. Look for **Send to Content Studio** on wallet events, token reports, vault snapshots, execution reports, and prediction-market cards - it prefills the generation surface with the on-chain context already attached. See [Wallet Intelligence](/thecontentforge-docs/web3-intelligence/wallet-intelligence.md), [Token Intelligence](/thecontentforge-docs/web3-intelligence/token-intelligence.md), [DeFi Intelligence](/thecontentforge-docs/web3-intelligence/defi-intelligence.md), and [Execution Intelligence](/thecontentforge-docs/web3-intelligence/execution-intelligence.md). ### How it uses Brand Voice Every generation pulls from your [Brand Voice](/thecontentforge-docs/feature-guides/brand-voice.md) profile: tone, vocabulary, length preferences, forbidden phrases, and example posts. The better your Brand Voice is tuned, the less editing drafts need. ### Web3 Project context For orgs configured as **Web3 Project**, Content Forge can pull from structured project context rather than relying on a loose prompt alone. That includes project profile, positioning, team, roadmap, FAQs and objections, approved supporting claims, and safe-to-use assets when available. See [Web3 Projects](/thecontentforge-docs/feature-guides/web3-project.md). ### Fast vs Best A pill toggle at the top of Content Forge picks the generation mode: * **Fast** - the quickest turnaround, good for high-volume work and iterations * **Best** - higher quality at the cost of extra seconds New users start on **Best** by default. Your choice is remembered per user. Generations kicked off from the Telegram bot always use Fast mode. ### Live progress While a job runs you'll see the current stage surface inline, such as "Gathering context", "Drafting", or "Refining", so you know it's moving. A **Context ready** chip appears near the submit button when your active mode's context is already loaded, so your first generation starts without a wait. ### Headline images For supported text modes, turn on **Generate headline images** to create platform-ready visuals alongside the copy. Image results can be downloaded, retried, or revised. Instagram square images are revised separately from the shared X/Facebook landscape image. Headline image generation produces **two takes at once**. When both return, you can toggle between them and keep the one you prefer, and a single failed take does not fail the request. (Image *remix* - regenerating from an existing image, e.g. a [Video Forge](/thecontentforge-docs/feature-guides/video-studio.md) thumbnail - returns a single image.) ### AI Revise Each output card has a **Revise** button. Open the inline panel, adjust the text, type a revision instruction ("shorter", "lead with the stat", "drop the question"), and Apply. Revise is quick and only updates that one platform's output. Links, @mentions, $cashtags, and factual claims are preserved unless your instruction explicitly asks otherwise. ### Create experiment from this draft Every output card has a **Create experiment from this draft** button. It opens the [Experiments](/thecontentforge-docs/feature-guides/experiments.md) builder with the draft prefilled as Variant A - fill in Variant B (and optionally C / D), pick a success metric, and you have a structured test instead of a guess. ### Generation Queue Multiple generations can run at the same time. The global [Generation Queue](/thecontentforge-docs/feature-guides/generation-queue.md) appears while jobs are active and shows each job's progress across the org. You can start a new generation without waiting for a previous one to finish. > **Note:** Comment King and Image Post return results inline and do not go through the queue. ### History and Recall Every completed generation is saved to History (last 15 entries). Each entry shows the mode, a summary of the input, a preview of the output, and a source badge if the generation came from the Telegram bot. Click **Recall** on any history entry to seed the input field with that generation's original input, letting you quickly iterate or regenerate with a different mode or platform. ### Publish Panel Every output card has a Publish Panel that lets you push, schedule, or manually log the post to a connected X account without leaving Content Forge (Instagram and Facebook publishing is coming soon). Attach media, pick a profile, choose **Post now**, **Schedule post**, or log something already posted, and submit. Scheduled and pending-approval posts show up on the [Calendar](/thecontentforge-docs/feature-guides/calendar.md). See [Publishing](/thecontentforge-docs/feature-guides/publishing.md) for the full flow. ### Workflow 1. Pick a mode and platform target 2. Choose Fast or Best 3. Provide the input: draft, link, topic, screenshot, document source, or image upload 4. Review the generated output 5. Revise inline, edit directly, retry images, or regenerate 6. Publish, schedule, manually log, or copy the final draft elsewhere ### Tips * If drafts feel "off," update Brand Voice rather than re-prompting every time * Use **Post Helper** for signal-driven posts; paste the topic or stat and let it write from your voice * Pair Content Forge with [Experiments](/thecontentforge-docs/feature-guides/experiments.md) when trying a new angle; don't guess whether it worked * Use **Recall** from History to resume a previous generation with a different mode or platform --- # 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-guides/content-studio.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.