跳到正文
cccskills
免费在 GitHub 公开

loop

Create and manage simple heartbeat automations attached to the current Codex thread. Use when Jason invokes $loop or asks this thread to keep going, check again, follow up, retry, monitor, or resume on a recurring cadence.

中文简介正在准备中。当前显示原文说明。

jxnl/personal-monorepo-template星标 573(GitHub 573 + cccskills 0)

实测评分 未实测

查看安装方法

包含的文件(1)

  • SKILL.md3.1 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Loop

Turn a plain-English request into a heartbeat on the current thread with as little ceremony as possible.

Set Up The Loop

  1. Infer the task from the request and current thread. Infer the stop condition when one is clear.
  2. Decide whether a heartbeat is useful. Use one when the work benefits from returning to this thread after time passes or external state may change; continue immediately instead when the task can simply be finished now.
  3. Use the requested cadence when sensible. Otherwise choose a cadence from the task's expected feedback latency, urgency, cost, and risk of noisy empty runs. Do not poll faster than the underlying state can plausibly change.
  4. Choose a short verb-led name and a self-contained task prompt. Put the work, its completion condition, and this terminal behavior in the prompt:
    • when the completion condition is met, rename the current thread to done: <short task name> using the thread-title tool without inventing a thread ID;
    • report the completed result and stop the loop. Keep timing and thread-target details in the automation fields.
  5. Call the automation tool with:
    • mode=create
    • kind=heartbeat
    • destination=thread
    • status=ACTIVE
    • the inferred name, prompt, and rrule
  6. After creation succeeds, rename the current thread to loop: <short task name> using the thread-title tool without inventing a thread ID.
  7. Return only the loop name, cadence, and what it will do.

Do not create a new thread, invent a thread ID, or substitute a detached cron job. Do not expose raw RRULE syntax to Jason.

Use Judgment

  • Ask when the task, cadence, stop condition, deadline, or notification behavior is materially ambiguous. Prefer the built-in user-question tool when it is available; otherwise ask one concise direct question.
  • When asking about cadence, offer a small set of task-appropriate choices and recommend one. Do not make Jason translate the task into scheduling jargon.
  • If the ambiguity is minor, make a reasonable assumption and state it instead of interrupting setup.
  • Treat “loop,” “keep going,” and “check again” as authorization to create the heartbeat requested in that message.
  • Preserve explicit deadlines, notification requirements, and stop conditions.
  • Keep the lifecycle prefixes exact: loop: while active and done: after successful completion.
  • Rename to done: only after successful completion, not after a failed run, pause, or temporary retry.
  • Prefer updating an existing matching automation over creating a duplicate.
  • For inspect, pause, resume, change, or delete requests, resolve the existing automation and use the corresponding automation-tool mode while preserving fields the user did not change. Restore the loop: title when resuming.
  • If the automation tool is unavailable, say so plainly. Never emit a raw automation directive as a workaround.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

assistant

免费

Meet or work with Assistant, the user's relaxed ongoing work support. Use when the user invokes Assistant, starts or resumes an Assistant chat, asks what they should know, wants proactive work awareness, asks for reply drafts, asks to keep an eye on work, or needs follow-up/check-in help. On first contact, start with exactly "Hi, I'm your assistant." then decide whether onboarding is brand new, partial, or already established.

中文简介正在准备中。当前显示原文说明。

jxnl/personal-monorepo-template5732026年6月23日 更新

Audit, de-slop, parameterize, modularize, or safely clean up AI-generated or AI-shaped backend/general code. Use for Python, TypeScript, or other implementation diffs that may contain duplicate helpers, fixture hacks, hard-coded test data, over-defensive control flow, broad exception wrappers, config-bag or boolean-mode soup, speculative scaffolding, hallucinated APIs/dependencies, local-idiom drift, brittle tests, or maintainability/safety/performance gaps after a feature, bugfix, prototype, or agent pass.

中文简介正在准备中。当前显示原文说明。

jxnl/personal-monorepo-template5732026年6月23日 更新

Audit AI-generated, AI-shaped, or AI-looking frontend code, UI screenshots, and design diffs. Use for prompts like "audit AI frontend", "de-slop UI", "componentize this screen", "parameterize this React/Tailwind/shadcn UI", "make it responsive/accessible", or "review design-system drift"; check component APIs, reusable props/data models, modular composition, shared primitives/tokens, responsive resilience, accessibility, copy quality, hard-coded fixture screens, one-off CSS piles, and generic cards/gradients/fonts.

中文简介正在准备中。当前显示原文说明。

jxnl/personal-monorepo-template5732026年6月23日 更新

Audit pasted chatbot output, AI-cleanup diffs, wiki drafts, Markdown/MDX/docs, and source-backed articles for generic AI fluff, LLM writing tells, weak audience model, lack of theory of mind, inflated significance, vague attribution, leaked tokens, placeholders, broken markup, fabricated or mismatched citations, and detector false positives. Use for requests like "AI writing audit", "check for AI slop", "find writing fluff", "does this sound like ChatGPT", "cleanup LLM tells", "make this less generic", "verify these citations", "detector flagged this", or text containing turn0search0, oaicite, oai_citation, contentReference, utm_source=chatgpt.com, malformed references, or wrong target-format markup.

中文简介正在准备中。当前显示原文说明。

jxnl/personal-monorepo-template5732026年6月23日 更新

Help address review/issue comments on the open GitHub PR for the current branch using `oai_gh` or `gh`; verify auth first and prompt the user to authenticate if not logged in.

中文简介正在准备中。当前显示原文说明。

jxnl/personal-monorepo-template5732026年6月23日 更新

gh-commit

免费

Review git changes and split them into semantic commits with clear messages. Use when the user asks to commit work, clean up local history, or group a mixed diff into logical commits. Do not commit on main or master unless the user explicitly asks.

中文简介正在准备中。当前显示原文说明。

jxnl/personal-monorepo-template5732026年6月23日 更新

查看 jxnl 的全部技能

报告此技能的问题