sip
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseTaste your own cooking: the moment you finish making something, check it with the very skills this repo ships before you serve it.
尝尝自己的“作品”:完成内容创作后,在交付之前先用本仓库提供的技能对其进行检查。
Goal
目标
A reminder buried in docs ("remember to verify") won't reliably fire in a fresh session. makes the recursive self-improvement loop a triggered habit: right after any create or change, run our own skills on the result — the clean checks and the true ones — so quality doesn't ride on the author's biased in-session judgment.
sip文档中提到的“记得验证”这类提醒,在新会话中往往无法可靠触发。 让递归式自我改进循环成为一个触发式习惯:在完成任何创建或修改操作后,用我们的技能对结果进行全面检查——包括基础检查和实质性验证——确保质量不受作者当前会话中偏见判断的影响。
sipWorkflow
工作流程
- Spot the trigger: you just created or changed an artifact or skill and are about to call it done, commit, or hand it off.
- Cold-read it — run on the artifact (fresh-eyes comprehension / handoff check).
shower - Verify it's true — if the artifact asserts a reality-grounded claim, run ; if it defines an eval, metric, or experiment, run
factchk. Skip when it has neither.mandela - Check consistency — run in audit mode across the repo for anything the change duplicated or contradicted; execute its consolidation plan only after approval.
ssotize - Detool portability claims — run only when the artifact claims portability, tool-neutrality, stack-agnostic durability, or cross-agent reuse. Skip when it is provenance, operational notes, a tool-targeted runbook, or does not claim portability.
detool - Tidy — the changed docs so the result reads as a clean v0, not a patch over a draft.
re0 - Apply the findings here, then serve it.
- 识别触发点:你刚创建或修改了一个工件(artifact)或Skill,正准备确定收尾、提交或交付。
- 客观审阅 —— 对工件运行 (以全新视角理解/交付检查)。
shower - 验证真实性 —— 如果工件包含基于事实的声明,运行 ;如果它定义了评估、指标或实验,运行
factchk。若两者都不涉及则跳过。mandela - 检查一致性 —— 在审计模式下对整个仓库运行 ,检查更改是否存在重复或矛盾内容;仅在获得批准后执行其整合方案。
ssotize - 验证可移植性声明 —— 仅当工件声称具备可移植性、工具中立性、跨栈持久化或跨Agent复用性时,才运行 。如果是溯源信息、操作说明、针对特定工具的运行手册,或未声明可移植性,则跳过。
detool - 整理优化 —— 对修改后的文档运行 ,确保最终内容是清晰的v0版本,而非草稿上的补丁。
re0 - 在此处应用检查结果,然后交付。
Rules
规则
- Trigger on your OWN output, right after making it — that's when bias is highest and a check is cheapest.
- Use the skills; don't re-implement them — for clarity,
shower/factchkfor truth,mandelafor SSOT,ssotizefor portability claims,detoolfor cleanup.re0orchestrates and routes findings back to the author session to fix; the skills do the work.sip - Skip what plainly doesn't apply, or any check whose skill isn't installed — run only what's present. /
factchkfire only when there is a claim or an eval;mandelafires only when the artifact claims portability, tool-neutrality, stack-agnostic durability, or cross-agent reuse; a one-line prose tweak may need only a consistency check. Say what you skipped and why.detool - Stop at the artifact — never touches git or makes commits.
sip - Chain only model-invoked skills; a user-invoked skill (marked ) is a human's to fire deliberately, so
disable-model-invocationmust not call one.sip
- 在你自己完成输出后立即触发——此时偏见最强,检查成本最低。
- 使用现有技能,不要重新实现——用于检查清晰度,
shower/factchk用于验证真实性,mandela用于维护SSOT(单一可信来源),ssotize用于验证可移植性声明,detool用于清理。re0负责编排并将检查结果反馈给作者会话以便修复;具体检查工作由各技能完成。sip - 明确跳过不适用的检查,或任何未安装对应技能的检查——仅运行已有的检查。/
factchk仅在存在声明或评估时触发;mandela仅在工件声称具备可移植性、工具中立性、跨栈持久化或跨Agent复用性时触发;一行文字的微调可能只需要一致性检查。请说明跳过的内容及原因。detool - 仅处理工件——绝不触碰git或执行提交操作。
sip - 仅调用模型可触发的技能;用户触发的技能(标记为)需由人工主动触发,因此
disable-model-invocation不得调用此类技能。sip
Verification
验证
Before finishing:
- The change was actually run through the relevant skills, not eyeballed.
- Findings were applied (or consciously deferred with a stated reason).
- The artifact actually changed as a result (a diff exists), or every skipped skill has a stated reason.
完成前需确认:
- 相关技能已实际运行于更改内容,而非仅靠人工目测。
- 检查结果已应用(或已明确说明延迟应用的理由)。
- 工件确实已做出更改(存在差异),或所有跳过的技能均已说明理由。