313 talks watchedbrowse the corpus

the proof · ai engineer 2026

The knowledge of AI Engineer 2026, on tap. Ask it something.

Your agent watched it — every sentence spoken, every line that crossed the screen, every frame. Every answer comes back with the sentence, the slide, and the second it happened.

1 result

Agentic Engineering: Working With AI, Not Just Using It — Brendan O'Leary

AI Engineer1 momentBEKc4P87XKo

5:15on-screenProblem 1: Context is expensive | Every token you add to the context window costs money - and the cost scales fast. | 10K | 200K | 600K | 1M+ | tokens | tokens | tokens | tokens | ~$0.03 | ~$0.60 | $1.50+ | More context ≠ better results | You're paying for noise, not signal | And even if cost weren't an issue - more context actually makes the model worse.youtu.be/BEKc4P87XKo?t=313 ↗

Add this corpus to your own agent

mcp endpoint

https://vidtheque.dev/mcp

claude code

$claude mcp add --transport http vidtheque https://vidtheque.dev/mcp

codex

$codex mcp add vidtheque --url https://vidtheque.dev/mcp

mistral vibe

$vibe mcp add vidtheque --url https://vidtheque.dev/mcp

Claude, ChatGPT and Vibe: add a custom connector and paste the endpoint. No sign-in.