메뉴
Use when starting any conversation — establishes vibekit's auto-trigger discipline so pipeline skills (brainstorm, brief, plan, exec, verify, review, finish, memory) actually fire at their trigger points instead of being silently skipped.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use after verify-gate returns `ready`, as an optional security pass over a vibe run's diff before finish-branch. Runs a tiered threat-model review — universal code-security checks on every diff, plus AI-artifact checks when the diff builds a skill, agent, prompt, or MCP server — quotes verbatim file:line evidence, scores findings, and hard-gates the handoff on CRITICAL/HIGH with a written-waiver escape. Peer to review-pack; both optional, either may run before finish-branch.
Use to diagnose vibekit installation health — skill file integrity, runtime registration parity, `.vibekit/` directory health, `docs/` subdir presence, authoring-contract alignment. Reports verdict only by default; `--fix` auto-repairs safe items. Cross-runtime portable.
Use after verify-gate returns `ready` and before finish-branch or any outward-facing action. Performs a self-review pass (Reflexion-style) — the skill critiques its own completed work against the spec and the plan, surfaces issues in order of severity, and presents the diff to the user for explicit sign-off. No outward-facing action runs until the user approves.
Use when the user says "vibe" or gives a short intent like "add X", "build Y", "fix Z". Runs the full disciplined pipeline end-to-end — brainstorm → plan → isolated execution → verification — with minimal user interaction outside the brainstorm phase. The user only has to answer design questions; every other step is autonomous, guardrail-protected, and token-efficient. If the user does not say "vibe" explicitly, do not assume this skill applies.
You MUST use this before any creative or implementation work — creating features, building components, adding functionality, or modifying behavior. Runs a disciplined Socratic brainstorming loop (hard gate, one-question-at-a-time, 2-3 approaches, design doc, user approval) while keeping token overhead minimal. Questions, user answers, approaches, design, and the written spec are never compressed.
Use when executing an approved implementation plan task-by-task. Dispatches one fresh subagent per task using an RTCO brief, receives the return through the report filter, runs two-stage review (self + plan-compliance), and only then marks the task complete and moves on. Prevents context pollution, drift across tasks, and silent regressions.
| name | using-vibekit |
| description | Use when starting any conversation — establishes vibekit's auto-trigger discipline so pipeline skills (brainstorm, brief, plan, exec, verify, review, finish, memory) actually fire at their trigger points instead of being silently skipped. |
This is not negotiable. You cannot rationalize "the task is too small," "I already know the answer," or "it would be faster to just do it." Skipping a guardrail skill is the failure mode this plugin exists to prevent.
If the user says "skip the brainstorm" or "just write the code," follow the user. Otherwise, follow the trigger map below.
| Trigger condition | Skill that MUST fire |
|---|---|
| User says "vibe" or gives a short intent ("add X", "build Y", "fix Z") | vibe |
| About to start any creative or implementation work, before code is written | brainstorm-lean |
| Spec is approved, implementation has not started | plan-write |
| Plan is approved, before any implementation dispatch | isolate |
| About to dispatch any subagent / Task / Agent call | brief-compiler (compile RTCO brief first) |
| Executing an approved plan task-by-task | exec-dispatch |
| Just received output from a subagent, tool chain, or dispatched task | report-filter |
| About to claim work is done, fixed, complete, or passing | verify-gate |
verify-gate returned ready, before any outward-facing action | review-pack |
| A vibe run is complete and the user wants a security pass on the diff (optional, before finish-branch) | security-review |
review-pack returned yes and user signed off on the diff | finish-branch |
| Driving a vibe run autonomously across iterations | ralph-loop |
| Storing, recalling, or auditing durable project knowledge | memory-dual |
verify-gate returns not satisfied/partial, or an exec-dispatch task fails its test/build | debug-recovery |
| User reports a bug, failing test, broken build, or asks "why is this failing" | debug-recovery |
| User reports vibekit feels broken, skills not firing, or cross-runtime drift | vibekit-doctor |
brainstorm-lean has a HARD-GATE; respect it on every project regardless of perceived simplicity.finish-branch is the only outward-facing endpoint.security-review is an optional peer gate. When the user runs it and it returns blocked (CRITICAL/HIGH finding), finish-branch is unavailable until the user fixes or waives — the same block-halts-handoff rule as review-pack.brief-compiler runs first; report-filter runs on return.Skill tool. The skill content loads — follow it directly. Never Read a skill file as a substitute for invoking it.AGENTS.md; invoke by following the named workflow.activate_skill tool.prompts/vibe.md is registered via the pi key in package.json; skills auto-discovered from skills/ and primed by the .pi-plugin extension. Invoke via /skill:<name> or by following the auto-trigger map.Compress agent chatter, briefs, capped reports, progress logs. Never compress questions to the user, user answers, plan files, TDD markers, verification evidence, error quotes, code blocks, destructive-operation warnings, or final commit/PR text. If a compression would reduce guardrail clarity by any amount, do not compress.
brainstorm-lean. The design can be three sentences, but it must exist and be approved.brief-compiler to produce an RTCO brief, every time.report-filter to enforce schema and strip restatements.