More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Turn Saved Videos into Searchable Records

    1

    Turn a pile of saved short videos into records you can search and use without ever opening the videos again.

    $12

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Turn Saved Videos into Searchable Records

    Example session with this skill installed

    Extract the recipe from this saved reel link. The video shows someone making pasta but doesn't state the boiling time or salt amount.

    • Read your context and instructions
    • Compiled the turn saved videos

    I have processed the video link and added it to the extraction queue. The transcription identified 4 sourced steps. Per the standing rules, missing values for boiling time and salt have been estimated based on standard pasta methods and flagged accordingly.

    Extracted Record: Garlic Butter Pasta

    FieldValue
    Sourcehttps://social-platform.com/reels/12345/
    StatusLive
    ProvenanceTranscribed / Standard Method
    IngredientsPasta, Garlic (3 cloves), Butter (2 tbsp), Salt (estimated), Water
    Sourced Steps1. Mince the garlic finely. 2. Melt butter in a pan over medium heat. 3. Add garlic to butter and sauté until fragrant. 4. Toss cooked pasta into the garlic butter sauce.
    Estimated Steps5. Bring a pot of water to a boil (estimated). 6. Add salt to water (estimated). 7. Boil pasta for 8-10 minutes until al dente (estimated).
    MetadataDifficulty: Easy (derived); Total Time: 15m (estimated)

    Next steps

    1. Review the provenance field in the record-schema.json to verify the source-to-estimate ratio.
    2. Run the build command to regenerate the static site with this new entry.
    3. Perform a click-test on the live UI to ensure the "(estimated)" labels appear correctly next to the boiling time.
    1. Review the field in the record-schema.json

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Saved-video collections rot: hundreds of clips, no search, and everything useful trapped inside players. This skill is the pipeline I used to turn saved Facebook recipes and Instagram reels into a working, searchable site. Its one rule: the written record is the product; the video is evidence.

    Inside the package: the skill file with the eight pipeline stages and the queue system, an example record schema, and the full extraction rules. The rules that matter most: every record must be complete enough to act on without watching the source; a value the source doesn't state — a quantity, a time, a temperature — may be estimated from similar records, but it is marked "estimated" in the stored data and shown as "(estimated)" on screen, never passed off as a fact from the video; and below a floor of real content, the record ships thin and honest instead of padded with invented steps.

    A run goes like this: the agent collects the saved items from the source platform, fetches each one's media and caption, transcribes or summarizes it (the summary is stored separately and never pretends to be the transcript), and extracts only genuine instructional steps — promotional filler like "follow for more" is excluded by pattern. Records carry a provenance field saying exactly how their content was obtained, so a stranger can judge how much to trust each one. The work runs from one queue file, in order, in fixed batches, with every item tracked as queued, live, or skipped with a reason. After the site rebuilds, the agent click-tests the live pages like a user — opens records, runs searches — because a correct data file once rendered a dead site, and only clicking caught it. Progress is reported as counts: live, queued, skipped.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Social Media.

    Requires a minimum of three sourced steps to avoid fabrication. Login-gated items require manual retry. Does not perform social actions on source platforms.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 6 days ago

    Software
    Saved short-form video linksA signed-in platform session

    Permissions

    Write Files
    Read Files
    Browser
    Network Access
    Listed6 days ago

    What's inside

    Frequently Asked Questions

    This skill solves the 'digital graveyard' problem by extracting actual data from saved videos, turning them into a searchable database of written steps and materials so you never have to re-watch a video to find a specific detail.

    Unlike basic prompting, this skill uses a strict evidence floor that flags any inferred values as 'estimates' and skips videos without enough actionable content, ensuring your database is accurate rather than halluncinated.

    The purchase includes the full automation pipeline for transcription and extraction, the JSON schema for data storage, and the static site generator template for your searchable frontend.

    The skill requires a signed-in session for platforms with login-gated content and a minimum of three distinct steps within the video to successfully generate a structured record.

    The skill exports data into standardized JSON schemas, making it highly compatible with most headless CMS platforms, static site generators, and personal knowledge management tools.

    New in Social Media