A YouTube Research MCP for ChatGPT and Claude
luntrasul9 · reddit · 2026-07-15
The author developed an MCP (Model Context Protocol) service named Hookami, integrating a YouTube research layer into ChatGPT and Claude.
The tool offers the following core features:
- Data Acquisition: Fetches trending keywords, keyword momentum, trend timing, driver videos, channel outliers, related topics, and the latest news.
- Security & Updates: Adopts a read-only mode, uses OAuth authorization, and refreshes the YouTube data pool daily.
The author observed different tool-calling habits across models: Claude tends to deeply call multiple tools before answering, while ChatGPT prefers to grab a few signals and move forward quickly. He is currently considering best practices for MCP tool design, such as whether to provide a fine-grained combination of small tools or a large tool capable of completing an entire workflow, and how much raw data tools should return to avoid context bloat.
More from coding & agent
- A better path to agent autonomy is running waves, finding friction, and iterating — JnBrymn · 2026-07-22
- Coding agents are heading toward an AI-writes, AI-reviews, human-approves workflow — aftahi_ai · 2026-07-22
- oMLX 0.5.2 adds Mac menu-bar stats, low-bit decode kernels, and faster downloads — awnihannun · 2026-07-22
- GitHub review bot hits its PR limit and forces a 39-minute cooldown — DanielLockyer · 2026-07-22
- Max reasoning effort appears to be mobile-only in Codex Remote, not desktop — GabGarrett · 2026-07-22
- A Reddit demo argues online stores should expose carts and pricing through MCP — gelembjuk · 2026-07-22