Once per repo, before anything else. The largest fixed load of any command, then it writes your .claude/.
Workflow slash commands for AI coding agents.
Guardrails, not autocomplete.
Supports Claude Code today. Cursor and Codex targets are planned.
Two runs, start to finish
Shipping a feature
Reviewing someone else's
Four commands
Each command triggers one distinct workflow/slashforge:setupOne-time repo setup. Explores the codebase, asks clarifying questions, then generates CLAUDE.md, rules, skills, agents, commands, and hooks tailored to it.
/slashforge:codeEnd-to-end development. Ten phases from plan to merged PR, with mandatory stop-and-confirm gates. Add -quick for lean mode on small changes.
/slashforge:investigateRead-only research. Reproduces and root-causes a bug, then writes a findings report. Never touches your code.
/slashforge:review-prReviews a pull request against your repo's own rules and conventions, then posts line-level comments or approves. Never posts without showing you the exact text.
What it costs
Tokens per runTen phases, four gates, agent code review. A multi-layer feature lands near the top.
The same run, indexed. The graph shaves roughly 4–10%; it does not change the order of magnitude.
Skips brainstorming and the review agent. Every user gate and all verification stay.
No code is written, so this is reading — however far the trail runs before the cause.
Almost entirely the diff. Three files is about 2k of it; fourteen is about 42k.
The range is driven by the size of the work, not by the tooling. SlashForge is deliberately heavy — if you want a prompt turned into a patch as fast as possible, this is the wrong tool.
It empowers people to do their finest work every single time.
Read the introduction