thenavidm/skill-builder
Skill creator plugin and skills for Claude Code, Codex and AI agents. Plans skills from a goal or from example repos, writes them to the Agent Skills standard, checks them in code, tests them with and without the skill, measures them in Claude Code, Codex, Gemini CLI and Cursor, and packages one source into every tool's plugin.
The entry point to Skill Builder, what it can do, which skill fits a request, setup and a health check. Use when unsure where to start with building or testing skills.
Runs a skill's evals in every installed AI tool, side by side. Use for "does it work in Codex or Gemini too", or before publishing to people on other tools.
Checks a skill or plugin against every AI tool's rules, in code, and explains each fix. Use when a tool won't load a skill, or before installing one someone sent.
Builds a better version of someone else's skill or plugin from scratch and proves it on the same tasks, with nothing copied. Use for "beat this skill". Not for testing your own (skill-eval).
Builds a new skill that does a task the user's way, from a goal, this chat or a brief, and proves it fires and helps before installing. Not for fixing one (skill-improve).
Checks that a skill's description fires on the right prompts and not on near misses, in each tool, and rewrites it until it does. Use when a skill won't fire or misfires.
Runs a skill's tasks with and without it in a real AI tool, grades every output and opens a review page. Use for "does this skill actually help", or to test one.
Gives every claim in a skill a source and a checked date, and removes advice nobody can source. Use before publishing a skill that gives advice or numbers.
Fixes a skill that gets something wrong, proven on the failing case and on everything it did right before. Not for tidying (skill-tidy) or new skills (skill-create).
Applies a correction to every skill it touches, with a diff to approve for each and their evals run again. Use for "from now on, never X" or after correcting an output.
Rebuilds every AI tool's plugin manifest (plugin.json and the rest) from one product.json and checks none has drifted. Use after adding or changing skills. Not for building a plugin (skill-plugin).
Plans a skill or plugin before building, deciding which jobs need a skill, an agent or a script, from a goal or example repos. Use for "plan a plugin for X".
Builds a whole plugin, or bundles existing skills into one, with an entry skill, and proves each request reaches the right skill in every tool. Not for rebuilding manifests (skill-package).
Puts a plugin repository on GitHub, private first, with a description and topics people search for. Runs only when asked by name.
Releases a new version that reaches every user, moving version, changelog and manifests together, then tag, GitHub and npm, each after a yes. Runs only when asked by name.
Sets up a plugin repository with its README as the spec, licence, changelog, CI, topics and one product.json for every tool's manifest. Use when starting a plugin repo.
Tidies a working skill that grew long, repetitive or stale while keeping its behaviour, proven by the same eval results at fewer tokens. Not for wrong output (skill-improve).
Measures what skills and plugins cost in context on every message, against each tool's budget. Use when skills stopped firing after installing more, or to slim a plugin.