Claude Code Mods — Development skill for Claude Code

claude-code-mods

Claude Mods and the tools to build them. The first tool is `mod-builder`, a skill that plans a mod's capability budget before it writes a line, then validates the result and writes the threat model. The first mod is `fable-pin`, which pins every subagent to one model by rewriting `agent.spawn`. The second is `cache-tax`, the Mod form of the hook of the same name in claude-code-hooks: it stops a cold send once with its price, keeps the 1-hour prompt cache alive for a window you set, and keeps score of what cold writes cost. Both were built with the skill and their READMEs carry the validator output and the threat model. All three install from this marketplace; the landing page is https://claude-code-mods.karanbansal.in/.

Install

claude plugin marketplace add karanb192/claude-code-mods
claude plugin install mod-builder@claude-code-mods
claude plugin install fable-pin@claude-code-mods
claude plugin install cache-tax@claude-code-mods

Then say "what is a Mod", "find a Mod", "build a Mod", "brainstorm Mods", "migrate this hook", "review my Mod", "debug my Mod" or "publish my Mod" in a session, or type `/mod-builder`.

What a mod is

A Claude Mod is a Claude Code plugin whose `hooks/hooks.json` names a TypeScript module. The module exports `register(on)`, and each hook is `on("event", matcher, async ($, e, next) => result)`. `$` is the engine interface, an object of nouns with verbs on each (`$.ui.log`, `$.fs.write`, `$.process.run`, `$.http.fetch`, `$.model.classify`, `$.agent.spawn`). `e` is the event. `next(e)` runs every hook beneath and then the engine, and returning without `next` answers in the engine's place. The code runs inside Claude Code's own process, with the process's reach, and costs no tokens unless it calls the model.

Anthropic calls the primitive function hooks. The design thread opened on 2026-09-03 and the built-in mods landed in the claude-code repo on 2026-09-09. It is early access. Nothing loads unless `CLAUDE