Demos written by your AI assistant.

Eight files: a SKILL.md, a system prompt, a JSON Schema, four worked examples, and a README. All publicly downloadable.

mkdir -p ~/.claude/skills/capturebeam/examples
cd ~/.claude/skills/capturebeam

curl -L https://capturebeam.com/agents/SKILL.md          -o SKILL.md
curl -L https://capturebeam.com/agents/system-prompt.md  -o system-prompt.md
curl -L https://capturebeam.com/agents/schema.json       -o schema.json

for n in 01-minimal 02-onboarding 03-feature-tour 04-storybook-component; do
  curl -L "https://capturebeam.com/agents/examples/$n.yaml" \
    -o "examples/$n.yaml"
done

1. Read the schema. Either fetch /api/v1/schema live, or read the static /agents/schema.json.
2. Probe the URL. Hit POST /api/v1/probe with { url }. The response lists every interactive element (role, name, label, hint).
3. Draft the YAML. Use NL targets like { role: "button", name: "Sign in" } — never CSS selectors. Keep demos short: 5–12 steps, ~30 seconds.
4. Render. POST to /api/v1/renders with raw YAML or a project ID. Poll /api/v1/renders/{id} every 2-3 seconds.
5. Self-correct. On failed, re-probe, identify the broken step from the per-step trace, patch, retry. Works without re-drafting the whole YAML.

Paste this into your agent's system prompt — or pull the current version directly from /agents/system-prompt.md.

You can produce demo videos using CaptureBeam.

Tools / endpoints (Bearer auth with the API key in $CAPTUREBEAM_KEY):
  GET  $BASE/api/v1/schema                 -> JSON Schema for demo.yaml
  POST $BASE/api/v1/probe   { url }        -> list of clickable elements
  POST $BASE/api/v1/renders { yaml }       -> { id, status: "pending" }
  GET  $BASE/api/v1/renders/{id}           -> { status, videoUrl?, error? }

To make a demo:
  1. Read the schema (live or from /agents/schema.json).
  2. Probe the deployed URL to identify real ARIA targets.
  3. Author YAML matching the schema. Use NL targets like
       { role: "button", name: "Sign in" }
       { role: "textbox", placeholder: "Email" }.
  4. POST the YAML to /api/v1/renders. Poll the job ID every 2-3s.
  5. On success, return the videoUrl. On failure, read the error,
     adjust the offending step (often: wrong target name or missing
     wait), and re-render.

Notes:
  - Each render is independent. No project setup required.
  - Steps run in order. A failed step doesn't fail the whole render —
    the runner records what it could and produces partial output.
  - Use `waitAfterMs` after navigations and form submits.
  - Add `caption` blocks to make the video self-explanatory.
  - render.quality: "1080p" (default, fast) | "1440p" | "4k" (3-5x slower).

From scratch — one shell script that drafts, renders, and prints the video URL. Replace $CAPTUREBEAM_KEY and the YAML body.

#!/usr/bin/env bash
set -euo pipefail

KEY=${CAPTUREBEAM_KEY:?}
BASE=${CAPTUREBEAM_BASE:-https://capturebeam.com}

YAML='title: Sign-up tour
subtitle: New user from scratch
render:
  quality: "1080p"
  preset: "midnight"
  aspect: "16:9"
  titleCard: true
steps:
  - type: goto
    url: https://app.example.com/signup
  - type: wait
    networkIdle: true
    ms: 1500
  - type: type
    target: { role: "textbox", placeholder: "Email" }
    text: "[email protected]"
    caption: { text: "Enter your email" }
  - type: type
    target: { role: "textbox", placeholder: "Password" }
    text: "supersecret"
  - type: highlight
    target: { role: "button", name: "Create account" }
    durationMs: 700
  - type: click
    target: { role: "button", name: "Create account" }
    waitAfterMs: 2000
    caption: { text: "And you are in." }'

JOB=$(curl -sS -X POST $BASE/api/v1/renders \
  -H "Authorization: Bearer $KEY" \
  -H "Content-Type: application/json" \
  -d "$(jq -nc --arg y "$YAML" '{yaml: $y}')" | jq -r .id)

echo "Render queued: $JOB"

while :; do
  R=$(curl -sS $BASE/api/v1/renders/$JOB -H "Authorization: Bearer $KEY")
  STATUS=$(echo "$R" | jq -r .status)
  case "$STATUS" in
    succeeded) echo "$R" | jq -r .videoUrl; break ;;
    failed)    echo "$R" | jq -r .error >&2; exit 1 ;;
    *)         sleep 3 ;;
  esac
done

Self-correcting agent

When a render fails on a missing target, the agent should probe the URL, find the closest matching element, and retry. The poll response includes a per-step steps array — index, type, status (ok/skipped/failed), optional error, and the resolved selector that was tried. That's enough for the agent to fix exactly the broken step instead of redrafting the whole YAML.

# Pseudo-code for a self-correcting agent
yaml = author_yaml_from_repo()
for attempt in range(3):
  job = post_renders(yaml)
  result = poll(job)        # blocks until done

  if result.status == "succeeded" \
     and not any(s.status == "failed" for s in result.steps or []):
    return result.videoUrl

  # Find the first failed step and ask: "what did the runner try?"
  first_fail = next(s for s in result.steps if s.status == "failed")
  print("step", first_fail.index, "failed:", first_fail.error)

  # Probe the page that step ran on, find a closest match by name.
  page = post_probe(deployed_url)
  fixed_target = best_match(first_fail, page.elements)

  yaml = patch_step(yaml, first_fail.index, fixed_target)

Per-PR demo bot

On every PR that touches a UI route, generate a demo video and comment with the embed. The agent reads the diff, identifies the changed page, and renders a flow that exercises the new code.

Onboarding videos from docs

For every "Getting started" markdown page in your docs, render a 30s walkthrough that demonstrates the steps. Re-render the whole set on a nightly cron — broken videos surface UI changes before users do.

• GET /api/v1/schema — JSON Schema + quick reference. No auth required.
• POST /api/v1/probe — discover elements at a URL. Bearer auth.
• POST /api/v1/renders — submit a render. Bearer + active subscription.
• GET /api/v1/renders/{id} — poll a render. Bearer auth.

Full reference + status codes: /docs/api.

• Concurrency: max 3 renders in flight per account. The 4th returns 429 — agents should back off or queue.
• Yearly safety net: 5,000 successful renders/year per account. Marketing copy says “unlimited fair use” — this is anti-abuse only. Email if you need it lifted.
• Probe: Bearer-authed but not subscription-gated, so an evaluating agent can use it before the user has subscribed.

• Synchronous render endpoint for short demos — saves the poll loop when an agent only cares about the result.
• Webhooks on render completion — for fully async flows (e.g. GitHub bots).
• Playwright-backed probe for SPA-heavy sites — current probe is static-HTML only.