FAQ

Which AI assistants work? Any agent that can read repository files and write content - Claude Code, Codex, Gemini CLI, IDE agents like Cursor, and similar. AGENTS.md is the cross-tool contract; CLAUDE.md simply points to it.

Can I paste the prompts into ChatGPT / claude.ai / a chat website? Not by itself. The prompts assume an agentic assistant launched in (or pointed at) the cloned Supercharger folder, because the output is files - tracks/<id>/…, preps/<id>/… - that the assistant must write to disk. A browser chat box can’t see your folder or save files into it. If a chat interface is all you have, you’d have to manually copy every generated file into the right paths per SPEC.md - possible, but the agent CLIs make this a one-line ask instead.

Does it cost anything to run? No. Supercharger has no API keys and makes no model calls - generation happens through whatever assistant subscription you already have. The app itself is free, MIT-licensed, and fully local.

Do I need Docker? Only for lessons with a terminal. Theory tracks (and the quiz/interview-prep portions of every track) work without Docker running.

How good is generated content, really? The agent contract enforces structure (scenario tickets, interview banks, tiered resources, closed-book finals) and the check scripts verify exercises actually validate. Substance still varies with the model and your prompt - iterate: tell the assistant what’s shallow and have it revise. The bundled SQL track is the calibration reference.

Does Supercharger trust AI-generated lessons? No. The assistant writes content against a spec, but sandbox lessons still need executable checks and human review. The strongest Supercharger tracks are not “AI said this is correct” artifacts; they are structured practice environments where the learner has to solve the ticket and pass validation.

Is this only for interview prep? No. Interview prep is the sharpest first use case because job postings provide clear requirements, but the same structure works for internal support enablement: reusable runbooks, practice tickets, onboarding tracks, and technical refreshers for recurring customer issues.

What is the Curriculum tab? It is the ordered study view for a prep. It combines existing tracks and requested gap tracks, shows lesson completion where a track already exists, and lets you queue missing tracks for an agent to create. When the track folder appears on disk, the button changes from a waiting state to Start.

Which confidence level should I choose? Choose Beginner when the topic or vocabulary is new, even if you are experienced in support. Choose Intermediate when you know the basics but need realistic ticket practice. Choose Advanced when you want edge cases, tradeoffs, and harder screens. The level tells the agent how much jargon it must define before expecting you to use it.

Why does Go deeper stop at advanced? Supercharger should help you build mastery, not create an infinite syllabus. After an advanced track, the better next step is usually practice, a final assessment, or a different adjacent skill rather than another deeper track on the same topic.

Can I request a track that isn’t tied to a job prep? Yes - the home screen’s Add track button records a standalone request in track-requests.json at the repo root. It shows as a dashed card among your tracks; Create queues it for your agent exactly like prep gap tracks. Track cards also show prep tags, so prep-owned and standalone tracks are distinguishable at a glance.

Does the Create button launch an AI agent? No. It records the request in track-requests.json and marks it creating. The GUI then shows a copyable prompt. Paste that prompt into Codex, Claude Code, Gemini CLI, or another assistant from the Supercharger project folder. This keeps Supercharger provider-neutral and avoids hiding credentials or process execution behind a browser button.

Ask your agent:

From the cloned Supercharger project folder, create the requested tracks marked creating in preps/<prep-id>/track-requests.json.
Follow SPEC.md and include createdBy and sourcePrep in each track.yaml.

When the assistant finishes, refresh the Curriculum tab. If tracks/<track-id>/track.yaml exists and is valid, the button changes to Start.

Does Add docs crawl an entire documentation site? No. It records approved docs URLs or pasted excerpts in onboarding-requests.json. The agent should use those approved sources to make product maps, glossaries, support scenarios, and product-specific track suggestions. This avoids bloated curricula from irrelevant docs pages.

Ask your agent:

From the cloned Supercharger project folder, create the product docs onboarding request marked creating in
preps/<prep-id>/onboarding-requests.json.
Use only the approved docs sources and flag unsupported assumptions.

Can I write tracks by hand? Yes - a track is just a spec-compliant folder. Hand-written, generated, or hybrid all render identically.

Is my data shared anywhere? No. No accounts, no telemetry, no network calls from the platform. Job postings you paste stay in preps/ on your disk. (Content you ask an assistant to generate is, of course, processed by that assistant’s model.)

Will updates overwrite my content or progress? No. progress.json and preps/ are personal state; tracks you’ve generated are yours. Pulling a new app version touches server/ and web/ only.

Why not just ask the assistant questions directly? You can - but structured curricula with validated hands-on labs, spaced drilling, and closed-book assessment is how familiarity becomes interview- ready competence. Supercharger turns one-off answers into a course you can finish and a claim you can defend.