More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Presenting Conference Talks: Slides and Script

    1

    Turn a paper into a talk people remember: one-sentence takeaway, story-spine slide plan with budgets per format, speaker notes, backup slides, and timed rehearsal discipline.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Presenting Conference Talks: Slides and Script

    Presenting Conference Talks: Slides and Script

    Example session with this skill installed

    My systems paper on a new caching layer got a 12-minute spotlight slot. I have 40 slides from my lab talks and I always run over. Turn this into a talk that fits and lands.

    • Read your context and instructions
    • Compiled the presenting conference talks

    ROOM FIRST: spotlight, 12 minutes on the program = 9 minutes of content once transitions and the session chair are priced in. One sentence the audience keeps: "Evict by predicted reuse distance, and tail latency falls fourfold." Every slide now has to serve that sentence or go.

    PLAN (from 40 lab slides to 10): 1) title with the sentence as subtitle (20s). 2) the problem — one production latency graph, headline: "p99 doubles at 70 percent load" (60s). 3) why the standard policy fails here — one diagram (60s). 4) the insight, alone on the slide (40s). 5–6) how it works — architecture diagram built in two steps (100s). 7) setup in one breath: two workloads, three baselines (30s). 8) the headline figure, full-width — tail latency down fourfold (70s). 9) one ablation: prediction is what matters, not the extra memory (50s). 10) summary + where to read more (30s). Total 8m 40s — inside budget with breathing room.

    CUT, NOT COMPRESSED: 30 lab slides removed, including the full related-work tour (now one positioning line on slide 3) and both extra workloads, which moved to backup with the overhead breakdown and the "why not a learned policy" answer — the three hardest questions, rehearsed.

    PRODUCTION AND REHEARSAL: Beamer chosen (equation-light but typography-consistent, no borrowed-laptop edits expected); speaker notes written per slide as transitions, not a script. First timed run: 10m 50s — over. Cut per the method (slide 9's second ablation panel, tightened openings) — second run 8m 55s, first 60 seconds rehearsed to automatic. Readiness rubric 9/10. The talk ends two minutes early and spends them in questions.

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

    About this skill

    A talk is not a paper read aloud — it is the advertisement for the paper, and this skill builds it that way. It starts with the room (audience mix, the true content minutes hiding inside your slot, and the one sentence the audience should remember a week later) before any slide exists. The story spine — why care, why it's hard, the insight, how it works, proof, what it means — is turned into a slide-by-slide plan with per-slide purpose and timing, sized by format budgets for poster talks, spotlights, orals, and invited talks. Design rules that survive hotel projectors (takeaway headlines, one idea per slide, figures over text, 24pt floor) come with a per-slide audit checklist, and the Beamer-vs-PPTX choice is made on explicit trade-offs. Speaker notes carry transitions, backup slides answer the three hardest questions, and rehearsal is a timed, out-loud discipline scored on a readiness rubric — including the late format change, the remote talk, and the demo that fails on stage.

    • It cannot rehearse for you: the timed, out-loud runs are yours to do, and the rubric scores them honestly only if you do.
    • Venue templates and accessibility rules vary; state them as constraints and the plan respects them, but it does not fetch your venue's kit.
    • If your headline results are still changing, result slides are built to swap — the method flags the risk but cannot freeze your experiments.

    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 any payment.

    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 24 days ago

    Permissions

    No special permissions declared or detected

    Listed24 days ago
    Updated21 days ago

    What's inside

    Frequently Asked Questions

    No — you start from the room and the one sentence, then audit what you have against the spine. Expect to keep the figures, lose most of the slides, and move the depth into backup. The demo shows a 40-slide deck becoming a 10-slide talk that way.

    The skill decides on explicit trade-offs: Beamer for math-heavy, version-controlled, consistent decks; PPTX when co-authors will edit at the last minute or media embedding matters. Either way, speaker notes and rehearsal discipline are the same.

    That is a named edge case: rebuild from the spine — problem, insight, headline result — rather than speaking faster through the long deck. The format budget table and cut rules are built for exactly that change.

    New in Research & Analysis