Command reference.
/hyperpower:build is the only command you need. It runs the other steps itself when it needs them. The rest are here for when you want to run one step on its own, check your setup, or look back at a past run.
Start here
Setup
- /hyperpower:buildStart here to build a feature or a whole app, or to take over an app you already have
- /hyperpower:helpEvery command, grouped by when you need it
- /hyperpower:initSet up this repo
- /hyperpower:doctorConfirm every check and tool still runs
- /hyperpower:configShow the settings in use, and where each one came from
- /hyperpower:pathsShow or set the folders it may edit
Run
- /hyperpower:runTake one requirement from start to finish
- /hyperpower:routeShow the steps a run would take, without running them
- /hyperpower:walkthroughExplain a pull request, or record your change for reviewers
- /hyperpower:resumeRerun a stopped run from any step
- /hyperpower:whyWhat a run decided, and what it assumed
- /hyperpower:correctFix an assumption a run got wrong
Memory
Suggestors
Analytics
- /hyperpower:usagePass rates, fix attempts and failed checks
- /hyperpower:costTokens and money by stage
- /hyperpower:telemetryControl what gets recorded
- /hyperpower:visualizeAn HTML report of what the agents did in a session
- /hyperpower:statusWhat hyperpowers is doing right now
- /hyperpower:councilAsk several agents one question, each answering on its own
- /hyperpower:standupThe same agents, replying to each other in rounds
Quality
- /hyperpower:humanizeFind and remove the marks a model leaves on code
- /hyperpower:sweepA model pass over the diff, for problems static tools miss
- /hyperpower:reviewSplit a diff into parts, review each part, then test every finding
- /hyperpower:respondAnswer every review comment against the code it points at
- /hyperpower:evalScore a prompt change
Checks and evals
29 commands.