See what models and agents do and cost, and learn to work with them.
- RelevePrice your Claude Code transcripts at real API rates: cost by model, project and skill, what the cache saved, and an editable rate card every figure recomputes from.
- Glass BoxClick through simulated Claude runs to see tokens, agent loops, subagents, MCP and planning mode, then drill scored practice questions.
- HeadroomPut 50 Kindle readers, six coding agents or a LoRa mesh on a DGX Spark, a Strix Halo or a Mac Studio, and find how many users it holds before you buy the box.
- Agent LoreLearn the agents, pick the model, ship the skill.
- DossierHand it the JSON an agent wrote and read the work: metric tiles with derived deltas, charts, before and after comparisons, and the evidence behind every claim.
- Prompt ForgeBuild character prompts for local LLMs and convert descriptions into image model tags.
- PrimerLearn ML, LLMs, and data insights by building small local projects with tested copy-paste code, then map each to the stack real companies run.