Three-Agent Audience Panel — EYE / PULSE / CROWD Verdicts Be

    1

    Three independent agents judge your finished video like an audience would — looks, entertainment, and click-worthiness — and it doesn't ship until all three say PASS.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Three-Agent Audience Panel — EYE / PULSE / CROWD Verdicts Be

    Example session with this skill installed

    Run the Audience Panel for 'how_to_bake_bread_final_v4.mp4'. Title: 'Why your bread is flat'. Thumbnail shows a deflated loaf. Context: Educational cooking channel.

    • Read your context and instructions
    • Compiled the three-agent audience panel

    I have convened the independent panel for your video. Each agent has reviewed the file and title package separately.

    Panel Verdicts

    Agent 1 — THE EYE

    • Verdict: PASS

    Evidence: 0:12 - Excellent lighting on the dough texture; 1:45 - Text overlays use high-contrast sans-serif that remains legible at 120px width.

    • Strength: The color grading on the final loaf is highly professional and fits the premium cooking niche.

    Agent 2 — THE PULSE

    • Verdict: FAIL

    Evidence: 0:02 - Hook is too slow; you are still setting up the camera. 2:15 to 3:10 - Significant drag while waiting for the mixer; no narration or b-roll here.

    • Fix List:
      1. Trim the first 3 seconds to start exactly when the bread is cut open.
      2. Cut the 55-second mixing sequence down to 10 seconds or add a voiceover explaining the science of gluten during this window.
      3. Increase voice energy at 4:20; the tone becomes flat during the yeast explanation.

    Agent 3 — THE CROWD

    • Verdict: PASS

    Evidence: The title 'Why your bread is flat' creates a strong curiosity gap that the video successfully resolves. 0:45 - The 'fail' examples are highly shareable.

    • Strength: The framing avoids all policy risks and fits the 'proven winner' format for home baking tutorials.

    Next steps

    1. Apply the two edits requested by THE PULSE to the master timeline.
    2. Re-export the file as 'how_to_bake_bread_final_v5.mp4'.
    3. Resubmit the new file for a full panel re-review; a FAIL from one agent requires a new pass from all three.

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

    About this skill

    A video can be flawless and still fail. The edit is clean, the facts are right, the export is perfect — and nobody clicks, nobody finishes, nobody shares. Technical checks can't catch that. Audiences can, but by the time they tell you, the video is already published.

    The Three-Agent Audience Panel puts the audience's judgment before the upload. Three independent agents — none of them your producer, none seeing each other's verdicts — judge the finished file from three angles:

    • THE EYE — does it look premium? Composition, color, phone-size legibility, polish, and an opening frame that survives the 120-pixel postage-stamp test.
    • THE PULSE — does it entertain? The 2-second hook, pacing, payoff, and ending pull, all judged with timestamps — plus the eight-year-old test (would a kid stay, or is this a lecture?) and the voice-energy check (narration must perform; flat is a fail).
    • THE CROWD — will it travel? Title and thumbnail honesty and pull, niche fit without copying, share-worthiness, correct audience targeting, and policy safety.

    Every verdict comes with timestamped evidence. Every FAIL comes with a numbered fix list concrete enough to execute without a meeting. And a PASS has to name what's genuinely strong — rubber stamps are treated as a panel malfunction, not a compliment.

    Run it after your technical QC and before you publish. Ship only on three passes.

    Buyer-facing benefits:

    • Get the audience's verdict before the audience gives it publicly — looks, retention, and click judged as three separate disciplines
    • Timestamped evidence for every judgment, so fixes are surgical instead of "make it better somehow"
    • The eight-year-old test kills lecture-style videos that accuracy alone would have waved through
    • Independence by design: three fresh judges, blind to each other and to the producer's opinion
    • Free — and the natural companion to the paid Final-Gate Video QC and Two-Agent Fact-Check Desk skills

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.

    Requires manual re-run of all three agents if any edits are made. Does not perform technical codec or file corruption checks. AI must have video/frame analysis capabilities enabled.

    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

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 14 days ago

    • Free with an Agensi account

    Permissions

    No special permissions declared or detected

    Listed14 days ago

    What's inside

    Frequently Asked Questions

    The skill requires an agent environment capable of processing video files and metadata, such as Claude (with file attachments) or Cursor. It functions as a structured workflow that directs three specialized agents to analyze your video from distinct perspectives.

    You receive the complete architectural framework for the three-agent panel, including the specialized personas for The Eye, The Pulse, and The Crowd, along with the standardized PASS/FAIL reporting protocol.

    Unlike simple prompts that often offer polite, vague feedback, this skill uses a multi-agent adversarial approach that forces the AI to provide timestamped evidence and conflicting critiques to ensure no pacing or visual flaws are missed.

    The skill is designed to analyze video content for any platform, including YouTube, TikTok, and Instagram, provided the agent environment supports the file format and length you are uploading.

    The panel provides a concrete 'fix list' with specific timestamps for every FAIL grade, allowing you to make targeted edits to pacing, lighting, or hooks before resubmitting for a final check.

    New in Agents & Orchestration

    Free