Specreel
Specreel turns recorded Playwright browser flows into narrated, hosted product demos that re-check themselves on a schedule. The CLI renders galleries from trace.zip files and republishes on every green build, and Specreel Cloud hosts galleries from a deployed URL with no CI setup.
Key facts
What is Specreel?
A demo and an end-to-end test become the same artifact, so your demo can't go stale without a test turning red.
Who is Specreel best for?
Developers and QA teams using Playwright, product/sales engineers, startups needing always-fresh product demos
What are its main limitations?
- The open-source CLI works primarily from Playwright traces, so a Playwright setup is needed unless you use recommend <url>
- AI narration is opt-in and requires bringing your own Anthropic API key
- Scheduled checks, alerts, private galleries and analytics are Cloud-only (Team/Scale), not in the free CLI
- The pricing page URL returned a 404 in the provided text, so plan details come from the homepage only
Limitations are based on product materials or editorial synthesis; confirm on the official site.
Key features
- ✓Renders a gallery of narrated demos from Playwright trace.zip test results (specreel test-results/)
- ✓Captions written from recorded clicks and assertions with no AI involved by default
- ✓GitHub Action (specreel/specreel@v1) republishes the gallery to the same URL on every passing build
- ✓Amber 'updated' badge for flows whose steps changed but still pass; broken flows turn the build red
- ✓--ai optional AI narration that rewrites captions into sales-engineer narration with your own Anthropic key, under a cent a flow
- ✓--bundle produces one self-contained gallery.html to email, publish to a URL, or embed with an iframe
- ✓specreel recommend <url> crawls a running app, ranks flows worth showing, and scaffolds a runnable Playwright script
- ✓--mp4 --voice exports an MP4 with real screenshots, step captions and a real voiceover track
- ✓specreel-mcp MCP server and Claude Code skill to run the whole crawl/pick/render/publish loop from an AI agent
Use cases
- →Generating and sharing a product demo gallery from an existing Playwright test suite
- →Keeping docs, sales and onboarding demos automatically in sync with the latest green build
- →Scaffolding Playwright tests from a running app when no test suite exists (specreel recommend <url>)
- →Emailing or embedding a self-contained demo as a single gallery.html file
- →Producing a narrated MP4 product tour with voiceover for prospects
- →Driving demo creation from Claude Code or Cursor via the MCP server by asking 'make a demo of my app'
Pricing
Self-hosted (Open Source, AGPL)
$0
- · Full CLI: render, gallery, MP4
- · recommend flow discovery
- · Single-file shareable bundle
- · Publish to GitHub Pages + embeds
- · GitHub Action (freshness loop)
- · MCP server & Claude Code skill
- · Runs on your machine and in your CI
Cloud
$0
- · 1 project, unlimited viewers, no card required
- · First 14 days include Team features, no card
Team
$39/mo
- · Unlimited projects
- · Scheduled checks
- · Alerts
- · Hosted galleries, no CI required
- · Stable per-project share URLs
- · Private galleries with access control
- · View & engagement analytics
- · Team workspaces & notifications
Scale
$149/mo
- · We run the browsers
- · Custom domain
CLI is free forever (open source, AGPL); Specreel Cloud is free for one project, with Team at $39/mo and Scale at $149/mo.
Pros
- +Open source under AGPL and free forever for the CLI, running locally and in CI
- +No AI required — captions are derived from recorded clicks and assertions
- +Demos cannot drift: they are re-published only on green builds, and broken flows fail the build
- +Python stdlib only, so no extra dependencies, and existing tests stay untouched
- +Flexible output: hosted gallery, GitHub Pages, single-file bundle, or MP4 with voiceover
- +MCP server and Claude Code skill let an AI agent run the entire loop
Cons
- −The open-source CLI works primarily from Playwright traces, so a Playwright setup is needed unless you use recommend <url>
- −AI narration is opt-in and requires bringing your own Anthropic API key
- −Scheduled checks, alerts, private galleries and analytics are Cloud-only (Team/Scale), not in the free CLI
- −The pricing page URL returned a 404 in the provided text, so plan details come from the homepage only
Positioning
- Core value: A demo and an end-to-end test become the same artifact, so your demo can't go stale without a test turning red.
- Ideal for: Developers and QA teams using Playwright, product/sales engineers, startups needing always-fresh product demos
- Product type: Open-source Python CLI (pip install specreel), GitHub Action, MCP server & Claude Code skill, and a hosted Cloud web app on Google Cloud Run
Frequently asked questions
What is Specreel?
Specreel makes a product demo and an end-to-end test the same artifact: it renders a recorded browser flow (Playwright trace.zip) into a narrated, hosted demo gallery that re-publishes on every green build.
How much does Specreel cost?
The self-hosted CLI is open source (AGPL) and free forever. Specreel Cloud is free for one project with unlimited viewers, the first 14 days include Team features with no card, Team is $39/mo and Scale is $149/mo.
Do I need an existing test suite?
No. With no tests yet, specreel recommend <url> crawls your running app, ranks the flows worth showing, and scaffolds a runnable Playwright script.
Does Specreel use AI?
By default captions are written from recorded clicks and assertions with no AI involved. AI narration is an opt-in --ai flag that rewrites captions into friendly sales-engineer narration using your own Anthropic key, for under a cent a flow.
How do the demos stay up to date?
A GitHub Action (specreel/specreel@v1) republishes the gallery to the same URL on every passing build, so a demo is never older than your last green build; a broken flow turns the build red instead of publishing a wrong demo.
Can an AI agent create the demo for me?
Yes. Specreel ships an MCP server and a Claude Code skill, so you can add it to Claude Code or Cursor and ask for a demo of your app to be made and published — the agent runs crawl, flow selection, render and publish on the same engine as the CLI.
You may also like
Curated suggestions based on similar categories.
Cloud Releases
Cloud Releases is the Google Cloud release notes documentation hub, where recent product changes across Google Cloud are published and indexed. The page offers an index of release notes, Google Cloud console release notes, and the ability to save and categorize content through collections. Google Cloud itself is offered with pay-as-you-go pricing, a pricing calculator, and free-tier options.
Diagramly
Diagramly is an AI-powered diagramming assistant that generates PlantUML, Mermaid, and ZenUML diagrams from written text descriptions or uploaded whiteboard drawings. Users can refine diagrams with diagram-as-code, save version history, and share diagrams via link or keep them private.
AdHelm CLI
AdHelm is a free, Apache-2.0 command-line tool that connects your AI agent to your own advertising and analytics accounts, so it can report performance, audit what was read and recommend changes entirely locally. It is read-only in version 0.1.101 and supports Google Ads, Meta Ads, LinkedIn Ads, TikTok Ads, Reddit Ads, OpenAI Ads, GA4 and GTM as EXPERIMENTAL providers.
OpenCode 2.0
OpenCode is an open source AI coding agent available as a terminal-based interface, desktop app, or web app. It supports many LLM providers, plugins, MCP servers, and custom commands, and can also be installed via Docker images.
IPTV Playlist
IPTV Playlist is a public M3U directory and browser-based utility set. It lists public IPTV playlist links organized by country, topic, language, and region from community sources, and lets users inspect, test, and analyze M3U streams.
AIAgentSkills
AIAgentSkills is a curated, community-driven directory for discovering AI agent skills (Claude skills) for Claude, Codex, and other AI agents. It publishes 8K+ searchable skills spanning categories such as writing, coding, design, marketing, testing, DevOps, security, and MCP integrations, each with a short description and install counts.