elan6666
UserYour own ByteDance, powered by Codex skills.
Categories
Indexed Skills (18)
byte-auto
Run Your ByteDance / Byte OS from idea to deliverable Byte Automatically. Use when the user asks for one-click completion, end-to-end Byte Delivery, auto mode, "do all steps", "don't stop until done", or to go from discussion to a shippable product with a ByteDance-inspired project team and evidence-led automatic iteration loops.
byte-brainstorm
Explicit-only Your ByteDance idea expansion mode. Use only when the user directly invokes `$byte-brainstorm` or explicitly asks to use byte-brainstorm. Expands rough ideas into multiple product directions, feature concepts, positioning angles, business models, and bold alternatives without entering the normal Byte OS workflow.
byte-build
Execute Your ByteDance / Byte OS plan files by dependency-ready waves. Use when the user asks to build, implement, execute plans, run the next wave, run `byte-build --wave N`, run `byte-build --plan ID`, or run `byte-build --all`.
byte-code-rules
Apply Your ByteDance engineering behavior rules for writing, reviewing, refactoring, planning, or executing code in Byte OS. Use with byte-build, byte-plan, byte-review, byte-iterate, byte-auto, or any coding task to avoid overengineering, keep edits surgical, surface risky assumptions, and define verifiable success criteria.
byte-codebase-harness
Prepare a large-codebase navigation harness for Your ByteDance / Byte OS that works across Claude Code and Codex. Use when the user asks to make a repo easier for Claude/Codex to navigate, add CLAUDE.md or AGENTS.md, configure large monorepos, map a codebase, scope tests/lint by directory, reduce generated-file noise, add LSP guidance, or support subagent exploration before editing.
byte-discuss
Discuss and clarify Your ByteDance / Byte OS product requirements before planning or building. Use when the user wants to discuss requirements, clarify an idea, ask what is unclear, confirm scope, compare product directions, or explicitly says not to write code yet.
byte-do
Route natural-language product work to the right Your ByteDance / Byte OS skill, decide whether goal mode or subagent mode is appropriate, and execute the matching workflow. Use when the user asks to start, discuss, continue, research, shape, plan, build, review, iterate, record a future idea, inspect status, analyze real users, auto-complete, parallelize, use subagents, or deliver a product using a ByteDance-inspired product squad.
byte-future
Record deferred Your ByteDance / Byte OS ideas that are explicitly not part of current execution. Use when the user says to save something for later, future plan, someday, parking lot, not now, 以后做, 未来计划, 先记下来, 暂时不做, or 下个阶段再考虑. Never turn parked items into plans, code, review blockers, auto-mode work, or delivery requirements unless the user explicitly promotes one.
byte-iterate
Run structured Your ByteDance / Byte OS iteration loops based on review findings, real user feedback, market research, test failures, metrics, experiments, or product goals. Use when the user asks to improve, optimize, iterate, create the next version, or run multiple automatic iterations.
byte-next
Continue a Your ByteDance / Byte OS project by inferring and executing the next workflow step from `.byte-os/` state. Use when the user says next, continue, resume, proceed, or asks what to do after the current stage.
byte-plan
Convert Your ByteDance / Byte OS specs into dependency-aware executable plans. Use when the product has been shaped and needs multiple plan files for design, engineering, testing, launch, OKR execution, or when the user asks to break the product into project plans before building.
byte-research
Research competitors, alternatives, pricing, market trends, user complaints, and product opportunities for Your ByteDance / Byte OS. Use when the user asks for competitor benchmarking, market intelligence, online research, product iteration based on external signals, or refreshed research for an existing `.byte-os/` project.
byte-review
Run a cross-functional Your ByteDance / Byte OS project review. Use when the user asks to review, audit, critique, check quality, simulate a ByteDance-inspired product meeting, assess readiness, or find issues before delivery or iteration.
byte-shape
Shape a Your ByteDance / Byte OS project into product positioning, MVP scope, non-goals, user flows, UX structure, technical direction, metrics, OKRs, and an initial roadmap. Use after byte-start or byte-research, or when the user asks to define what the product should be before implementation.
byte-start
Start a Your ByteDance / Byte OS project from a rough idea. Use when creating a product from zero, initializing `.byte-os/`, combining discovery and research into the first foundation step, or restarting early-stage product discovery with a ByteDance-inspired project team.
byte-status
Summarize Your ByteDance / Byte OS state, artifacts, OKRs, plan progress, current stage, blockers, and next recommended action. Use when the user asks for status, progress, current stage, what exists, what is missing, or where to continue.
byte-users
Analyze real post-build user feedback for Your ByteDance / Byte OS. Use only when a product, prototype, URL, screenshot, recording, analytics export, interview notes, survey results, support tickets, reviews, or other real user evidence is available. Do not use for simulated personas.
byte-deliver
Package final Your ByteDance / Byte OS delivery artifacts. Use when the product has been built and reviewed, or when the user asks to ship, hand off, summarize, package, or produce final delivery instructions.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.