Keeps a change scoped to what was asked. Flags extra features, refactors, and 'while we are here' edits, and moves them to a suggestions list instead of the diff. Use when a request is small and the agent is about to expand it, when a PRD grows beyond its problem statement, or when the user says 'just do what I asked'.
Wayne Sutton
Startup Programs at Convex + Head of Community & Events at @convex / Connecting developers & startups to build faster with Convex. / ๐ Adv Rider /
Mostlyยทconvex, TypeScript, validation
Skills
waynesutton/builder-skills
Workflow, security, and writing skills for builders shipping Convex apps. npx builder-skills
npx skilld add waynesutton/builder-skills --allRetired. Points to the skill that replaced convex-agents in builder-skills 3.0. Use when /convex-agents is invoked by name or the user asks where convex-agents went.
Retired. Points to the skill that replaced convex-best-practices in builder-skills 3.0. Use when /convex-best-practices is invoked by name or the user asks where convex-best-practices went.
Retired. Points to the skill that replaced convex-component-authoring in builder-skills 3.0. Use when /convex-component-authoring is invoked by name or the user asks where convex-component-authoring went.
Retired. Points to the skill that replaced convex-cron-jobs in builder-skills 3.0. Use when /convex-cron-jobs is invoked by name or the user asks where convex-cron-jobs went.
Retired. Points to the skill that replaced convex-file-storage in builder-skills 3.0. Use when /convex-file-storage is invoked by name or the user asks where convex-file-storage went.
Retired. Points to the skill that replaced convex-functions in builder-skills 3.0. Use when /convex-functions is invoked by name or the user asks where convex-functions went.
Retired. Points to the skill that replaced convex-http-actions in builder-skills 3.0. Use when /convex-http-actions is invoked by name or the user asks where convex-http-actions went.
Retired. Points to the skill that replaced convex-migrations in builder-skills 3.0. Use when /convex-migrations is invoked by name or the user asks where convex-migrations went.
Retired. Points to the skill that replaced convex-realtime in builder-skills 3.0. Use when /convex-realtime is invoked by name or the user asks where convex-realtime went.
Retired. Points to the skill that replaced convex-schema-validator in builder-skills 3.0. Use when /convex-schema-validator is invoked by name or the user asks where convex-schema-validator went.
Retired. Points to the skill that replaced convex-security-audit in builder-skills 3.0. Use when /convex-security-audit is invoked by name or the user asks where convex-security-audit went.
Retired. Points to the skill that replaced convex-security-check in builder-skills 3.0. Use when /convex-security-check is invoked by name or the user asks where convex-security-check went.
Writes a short PRD in prds/<slug>.md before non trivial work: problem, root cause for bugs, proposed solution, files to change, edge cases, verification steps, and a completion log, with UTC metadata. Then queues the work in task.md. Use before any multi file feature, schema change, architectural decision, or bug that took real investigation, or when the user says write a PRD, @create-prd, or plan this.
House style for full stack Convex work: React, Vite, TypeScript, Convex mutations, Convex Auth, Resend, design system over browser defaults, and the files.md, changelog.md, task.md habit. Edit it to match your project after installing. Use when building features, writing Convex functions, or making code changes in a project that uses this style.
Stops an agent from destroying uncommitted work with git. Requires git status and git diff before any command that discards changes, treats "revert" and "undo" as manual edits, and never runs reset, checkout, clean, stash drop, or force push without explicit approval. Use when a task involves reverting, undoing, resetting, cleaning, rebasing, or force pushing, or before any git command in a repo with uncommitted changes.
Security check and audit for Convex apps and any app with a React, Vite, Next.js, or Expo frontend, built from real incidents across shipped repos. Runs a secret scanner plus a per function surface ledger, then gates every candidate finding on a real attacker path before it gets a severity. Covers leaked keys, client prefixed secrets, public functions with no auth, caller identity taken from args, paid AI and AI Gateway spend, unsigned webhooks, open uploads, unsafe HTML, prompt injection into agent tools, component boundaries, derived data that outlives its source, status moves that skip steps, unbounded reads, and CI deploy keys. Use before any deploy, before making a repo public, when adding AI, webhooks, uploads, env vars, admin features, or agent tools, when reviewing AI generated code or a PR, when checking whether a security fix worked, or when the user says sec check, security check, security audit, pen test, find vulnerabilities, verify the fix, or is this safe to ship.
Finds the implicit state machines in an app: status, state, stage, phase, and step fields, groups of booleans that act like one status, and flows spread across UI, mutations, webhooks, and scheduled jobs. Reports every state, every function that moves it, who can trigger each move, and the problems, then proposes an explicit transition map. Report only by default. Use when the user says find the implicit state machines, audit status fields, map the order or onboarding flow, or when a status bug, a skipped step, or a double run refund, payout, or email shows up.
Reflection first problem solving for Convex projects: root cause before code, two candidate fixes, a 98 percent confidence bar, and changes limited to the files the request needs. Use before implementing any fix or feature, when a bug report comes in, when a previous fix did not work, or when the request is vague.
Brings task.md, changelog.md, and files.md back in sync with what the code says shipped, then prints a commit message as plain text. Reads git history and the working tree, never memory, and refuses to log features it cannot see in source. Use when a feature or fix lands, when the user says /update, @update, update project docs, sync changelog, or update files.md, or when task.md and the code disagree.
Ends every agent turn with a clear "What I need from you" section listing concrete actions the user must take, or says "Nothing needed from you right now", and logs every ask by id and session in prds/what-i-need-from-you.md. Use on every turn, especially when blocked on the user, when the next step needs a login, dashboard, key rotation, decision, file, or approval, or when finishing a task, audit, or fix.
Runs multi step work the way a careful senior engineer would: triage before code, a PRD in prds/, task.md kept current, docs synced after, and lessons recorded when a mistake repeats. Hands PRD writing to create-prd and the docs sync to update-project-docs. Use when a task spans several files or steps, changes architecture, needs a plan the user can review, or when the user says plan this, track this in task.md, or what did we learn.
Drafts, proofs, and audits prose in the author's voice, matched to the reader and destination. Covers social posts, articles, emails, docs, READMEs, commit messages, and feature announcements, with evidence rules that forbid invented facts, quotes, or experiences, and a draft only default that never publishes or sends. Use when drafting, rewriting, proofreading, or auditing social posts, articles, emails, docs, READMEs, or commit messages, or when the user says write:, proof this, or make this sound less like AI.
Curated skills from @waynesutton. Checked GitHub just now