What is ReelQL and who builds it?

A Claude skill plus API from tomascupr that takes any video link and returns one typed JSON document describing the video: summary, chapters, scenes, cast, key moments, brands, on-screen text, and transcript.

How does ReelQL use Jev?

Its scripts/jev.py asks Jev five questions per video: on brief, hook, product in use, emotional payoff, and unsafe content. The README says Jev returns probabilities your code can branch on, but the digest does not show the full request shape.

How do I start using ReelQL?

Install the Claude skill from the repository and set both keys, ReelQL's and Jev's, then ask Claude to analyze or rank videos. The digest shows no install command for the skill itself, but two demo GIFs walk through real runs.

What does ReelQL cost?

The README does not state pricing in the captured text. It shows per-video analysis times from 16 seconds for a 6-second TikTok up to 65 seconds for a 14-minute walking tour, and says the workflow needs both keys set.

Can I see sample output without a key?

Yes. The README points to the examples/ folder, which holds the full output for five brand TikToks plus the Jev ranking of them, for anyone who has no key yet.

Does ReelQL do the ranking itself?

No. ReelQL produces the typed analysis; Jev makes the judgments. In the demo, the skill runs every link through ReelQL and then asks Jev five questions per video, so the ranking numbers are Jev's probabilities.

What is missing from the README capture?

The digest truncates the Python example mid-way through building the state dict, so full typesafe_sdk usage, error handling, pricing, and rate limits are not documented here. Metadata shows 32 stars, HTML as the primary language, MIT, created 2026-09-25.

Jev directory

CommunityJev Agents & MCP

ReelQL

A Claude skill and API from tomascupr that turns any video link into one typed JSON document: summary, chapters, scenes, cast, key moments, the emotional arc, brands, on-screen text, and the full transcript. Jev then judges those fields so your code can branch on probabilities.

See highlights Open github.com

Stars: 32
Language: HTML
License: MIT
Reading time: 2 min
Published by: Community
Author: tomascupr
Last verified: Oct 2026

tomascupr/reelql

User repository on GitHub

View on GitHub

Give your agent eyes: any video link in, one typed JSON out. A Claude skill + API.

Forks: 1
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.

Tagscommunity github claude video python

Works forJev for Extraction Jev for Scoring Jev for Ranking

Used inJev for Media & Entertainment Jev for Sales & Marketing

Languages: Python
Platforms: API
Capabilities: Open sourceRequires API key

Highlights

A Claude skill plus API: paste any video link and get one typed JSON document back with every field timestamped.
Output covers summary, chapters, scenes, cast, key moments, emotional arc, products and brands, on-screen text, and the full transcript.
README speed table: a 6-second Starbucks TikTok analyzed in 16 s, a 4-minute unboxing in 23 s, and a 14:05 walking tour in 65 s.
Its bundled scripts/jev.py asks Jev five questions per video: on brief, hook, product in use, emotional payoff, and unsafe content.
Ranking demo: five brand TikToks judged against "family-friendly, shows the product in use"; Samsung scored 77% and Nat Geo 6%.

Watch out

README and GitHub metadata reviewed 2026-10-17; the digest truncates the bundled Python example that imports typesafe_sdk, so the full code, pricing, and rate limits were not available. The latency table and the 77% / 6% ranking are the author's own recorded runs on a handful of videos, not an independent benchmark.

Questions about this entry

Short answers drawn from the linked project or page.

From the Jev guides

Original write-ups on what this entry is for.

MCP tool routing: Jev in front of the tool call

Oct 16, 2026

Tool menus can outgrow the context that reads them. The catalog's Jev-based routers and gates narrow the list: benchmarked tool search, per-tool scoring, fail-closed approvals, and WebMCP in the browser.

Jev Agents & MCP#mcp#tool-routing#agents

Read article

Approval policies for coding agents: hooks, guards, and gateways