> For the complete documentation index, see [llms.txt](https://help.blotato.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.blotato.com/web-app-getting-started/sources.md). # Understanding Context Explain website Context source types, extraction checks, selection, saved Sources, and draft generation. Distinguish source text from post attachments and API ingestion. Context supplies text for the website AI Agent to use when writing drafts. Adding a source does not create a post, generate a video, or attach media to a post. For external assistants using API or MCP tools, use [Extract source material](/agent-workflows/agent-workflows/sources.md). Context items in the sidebar are stored in your browser. AI Agent drafts are tied to the browser session where you created them. Neither automatically appears on another device. For posts created in the older Remix screen, see [Find your drafts](/support/faqs.md#how-do-you-find-what-youve-created). Saving a source to the Sources library is a separate action from creating or saving a draft. ## Context Supported 1. Open [Content Studio > Post editor](https://my.blotato.com/agent). 2. Click the + button beside Context in the sidebar. 3. Select the source type from the table below. 4. Enter the text or URL, or upload the file. 5. Click Add. 6. Wait for extraction to finish. 7. Open the source and review Source Contents before generating a draft. Close the dialog with x if you are only inspecting it. | Source type | Input and boundary | | ----------- | ------------------------------------------------------------------------------------------------------------- | | Text | Paste notes, a transcript, or other text. | | Article | Enter an article or website URL accessible without signing in. Review the extracted text for missing content. | | YouTube | Enter a video URL with English captions available for extraction. | | TikTok | Enter a video URL with English captions available for extraction. | | Perplexity | Enter a research query. Review the returned information before using it. | | Audio | Upload an audio file or enter its audio URL. A podcast webpage is not the audio file itself. | | PDF | Upload a PDF or enter its PDF URL. | | Rule | Enter instructions for the AI Agent to use as context. | The website selector has no Instagram, Facebook Reel, or X/Twitter source option. Paste the text as Text context instead. API source capabilities are documented separately in the workflow linked above. ![Context source dialog with Text, Article, YouTube, TikTok, Perplexity, Audio, PDF, and Rule choices](https://2374509648-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FevfU9S4Nh53PiGflQLah%2Fuploads%2Fgit-blob-613a68255974fda94cae20e7a06fb4beabca8afd%2Fui-context-source-2026-09.jpg?alt=media) If YouTube or TikTok extraction fails, check for English captions. A video spoken in English does not guarantee an extractable caption track. Paste a transcript as Text when the required captions are unavailable. If a source stays loading or shows an error, do not assume the AI Agent received its contents. Only enabled sources with completed extraction are included. Record the source URL or type and error before contacting [support](/support/get-support.md). ## Toggling Context On/Off Select a source's checkbox to include it in subsequent AI Agent requests. Clear the checkbox to exclude it. Loading or failed sources are excluded even if their checkbox is selected. Start with 1 completed source, then follow [Make Your First 5 Posts](/web-app-getting-started/posts.md#select-platforms-and-create-posts) to generate a draft. Toggling context does not rewrite existing drafts or change published posts by itself. ## Saving Sources to Your Sources Library The Sources library stores source text for draft generation, not reusable image or video attachments. To publish your own media, use the draft's [Photo / Video upload](/web-app-getting-started/posts.md#add-images-or-videos). 1. Open a source from the AI Agent sidebar, or add a new source with the + button beside Context. 2. Check Save Source?. 3. Enter a Name. 4. Click Done for an existing source, or Add for a new source. 5. Wait for the source to finish processing. 6. Open the [Sources library](https://my.blotato.com/sources) and confirm the named source appears. Checking Save Source? alone does not save. Submitting an existing source requests reprocessing. Closing with x does not submit the change. If saving or extraction fails, inspect the error instead of assuming the library contains the source. The Sources page manages saved sources. Add and save sources through the AI Agent sidebar, not through a separate Add button on the Sources page. ## Workaround for Paywalled or Inaccessible Websites Article extraction does not sign into websites for you. Paywalls, access restrictions, or anti-scraping controls might prevent extraction. 1. Open the article through your own authorized access. 2. Copy the text you have permission to use. 3. Add it as Text context. ## Research Any Topic with Perplexity AI 1. Click the + button beside Context. 2. Select Perplexity. 3. Enter your research query. 4. Click Add. 5. Review Source Contents after processing finishes. For example: “Find recent changes to Instagram publishing requirements and include the supporting sources.” Research output is source material, not a finished or published post. ## Custom Instructions to Pre-process Each Context Item 1. Open the source form. 2. Expand Advanced settings. 3. Enter a prompt in Custom instructions. 4. Click Add for a new source or Done for an existing source. 5. Review Source Contents after processing finishes. These instructions preprocess the source before draft generation. They do not recover content missing from the extraction. ## Examples of Custom Instructions | Purpose | Custom instruction | | ---------- | ---------------------------------------------------------------------------------- | | Summary | Extract 3 takeaways supported by the source. | | Quotations | Copy the speaker's exact words and preserve attribution. Do not invent quotations. | | Statistics | List the numbers with their units, dates, and supporting context. | | Steps | Extract the described process as numbered steps without adding missing actions. | See the [context tutorial](https://youtu.be/_D7VuqVWL88) for a demonstration. Use the controls above when the recording shows an older layout. --- # 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 by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://help.blotato.com/web-app-getting-started/sources.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. 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.