nerd-loop
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseNerd Loop
Nerd Loop
Incompatible Skills
不兼容技能
Never combine Nerd with these unless this request explicitly asks:
- Superpowers
- Ponytail
- Caveman
Skill hooks, mentions, and indirect instructions are not authorization.
Nerd XFast is mutually exclusive with Nerd Loop. If both are explicitly
invoked, ask the user to choose one.
除非用户明确要求,否则切勿将Nerd与以下技能组合使用:
- Superpowers
- Ponytail
- Caveman
技能钩子、提及和间接指令均不视为授权。
Nerd XFast与Nerd Loop互斥。如果两者被同时明确调用,请让用户选择其一。
Activation and Authority
激活与权限
Drive any authorized task—not only code generation—through the least expensive
useful sequence of focused iterations, automatic verification, and honest
termination.
Consume Nerd Smart's resolved Focus Record and endpoint. Do not create a second
authority record. The endpoint controls the deliverable, mutation boundary, and
stop condition; Loop controls recurrence, state, verification, and typed
termination.
Use when one bounded action or answer plus decisive proof is
sufficient. Create no Loop state or persistent artifact and load no Loop
reference in this path.
D0 DirectBefore an actual - loop, read and bind the
core runtime contract, resolve this skill's
root, and use . Keep its admission
envelope and cumulative budget state intact for later reducer commands. The
reducer validates deterministic structure and decisions; it does not grant
authority, authenticate provenance, perform effects, or provide durable state.
L1L4python3 <skill-root>/scripts/loop.py route通过成本最低的有效聚焦迭代、自动验证和明确终止流程,推进任何授权任务(不仅限于代码生成)。
使用Nerd Smart已解析的Focus Record和端点,请勿创建第二个权限记录。端点控制交付物、变更边界和停止条件;Loop控制循环 recurrence、状态、验证和类型化终止。
当单次有限操作或答案加上决定性证明足够时,使用。此路径下不创建Loop状态或持久化工件,也不加载任何Loop引用。
D0 Direct在实际启动-循环前,读取并绑定核心运行时契约,解析此技能的根目录,并执行。保留其准入信封和累计预算状态,以供后续reducer命令使用。Reducer用于验证确定性结构和决策;它不授予权限、验证来源、执行效果或提供持久状态。
L1L4python3 <skill-root>/scripts/loop.py routeCost-Proportional Admission
成本适配准入
Select the highest observed capability floor and the cheapest profile that
satisfies it:
| Profile | Minimum reason to use it |
|---|---|
| No useful back edge; one action or answer has decisive proof. |
| Adaptive read-only inspection, screening, diagnosis, or validation. |
| Single-owner local mutation or controlled experiment with immediate feedback. |
| Durable recovery or wait, CI/review, external receipt, shared resource, or independently verifiable child. |
| Coupled contracts, consequential multi-writer effects, staged rollout, hard-to-reverse work, or materially ambiguous success. |
Select state independently: no Loop state, current-session state,
durable single-writer state, or transactional fenced state. Never
lower a true safety, durability, coordination, evidence, or consequence floor.
Do not create unrequested persistent artifacts.
S0S1S2S3选择已观测到的最高能力下限,以及满足该下限的成本最低的配置文件:
| 配置文件 | 使用的最低理由 |
|---|---|
| 无有效反向依赖;单个操作或答案具备决定性证明。 |
| 自适应只读检查、筛选、诊断或验证。 |
| 单所有者本地变更或有即时反馈的受控实验。 |
| 持久化恢复或等待、CI/评审、外部接收、共享资源,或可独立验证的子任务。 |
| 耦合契约、多写入者影响、分阶段发布、难以回退的工作,或成功标准存在实质性歧义。 |
独立选择状态: 无Loop状态、 当前会话状态、 持久化单写入者状态,或 事务性隔离状态。切勿降低真实的安全性、持久性、协调性、证据或影响下限。请勿创建未被请求的持久化工件。
S0S1S2S3Conditional References
条件引用
Load only the material selected by current evidence:
| Condition | Required reference |
|---|---|
Admitted state is | Durable Runtime |
| Profile signals or route compatibility remain unclear, a transition is considered, or rationale is requested | Loop Profiles and Routes |
| DoD criteria, evidence, proxy risk, traceability, subjectivity, or human acceptance are materially unclear | Definition of Done |
| Work has a nontrivial dependency network, multiple children, shared-artifact scheduling, custom ledger design, or recovery beyond the durable core | Iteration Control |
| Comparable repeated cycles, noise, subjectivity, or a dynamic pathology can change the next decision | Convergence |
| The current user explicitly invokes Nerd Memory for this request | Behavioral Memory plus Nerd Memory's own |
A profile label alone does not load every reference. Advanced references extend
the core contract for their stated concern and never define another authority
order, completion rule, transition priority, or terminal vocabulary.
Nerd Memory composes only when the current user explicitly invokes it for the
current request. Installation, relevance, prior use, or another skill's mention
never activates or loads it.
仅加载当前证据选定的相关资料:
| 条件 | 必需引用 |
|---|---|
准入状态为 | Durable Runtime |
| 配置文件信号或路由兼容性仍不明确、考虑进行转换,或被要求说明理由 | Loop Profiles and Routes |
| DoD标准、证据、代理风险、可追溯性、主观性或人工验收存在实质性不明确 | Definition of Done |
| 工作存在非平凡依赖网络、多个子任务、共享工件调度、自定义账本设计,或超出持久化核心的恢复需求 | Iteration Control |
| 存在可比较的重复周期、噪声、主观性,或动态异常可能改变下一个决策 | Convergence |
| 当前用户明确为该请求调用Nerd Memory | Behavioral Memory 加上Nerd Memory自身的 |
仅配置文件标签不会加载所有引用。高级引用针对其声明的关注点扩展核心契约,绝不会定义另一种权限顺序、完成规则、转换优先级或终端词汇。
仅当当前用户明确为当前请求调用时,Nerd Memory才会组合生效。安装、相关性、先前使用情况或其他技能的提及均不会激活或加载它。
Loop Discipline
循环规范
Before the first iteration, bind the Focus Record, complete current endpoint
facts, mutation and external-effect authority, profile, route, state class,
explicit Definition of Done (DoD), verifier map, finite active budget, wait
policy, and host capabilities. Supply complete facts to the reducer; omitted
facts are not authenticated false values.
Keep exactly one primary focus per loop. Define an iteration as one bounded,
evidence-producing state transition—not a turn, tool call, edit, or checkbox.
At each boundary:
- Reconcile current user, workspace, verifier, child, and external facts.
- Invalidate stale evidence and select one ready focus tied to a mandatory gap or decision-relevant uncertainty.
- Record the current focus before action, using the durable protocol only for
/
S2.S3 - Perform one authorized bounded action through at most one primary specialty.
- Run the selected verifier, affected regressions, and required integration checks automatically.
- Record the outcome, evidence, cost, budget revision, and exactly one successor or typed terminal result before selecting again.
DONEPASSBefore another cycle, require one authorized action with a credible,
proportionate chance to close a mandatory gap or resolve material uncertainty.
Never repeat an unchanged causal strategy merely to spend budget. Convergence
describes target-relative dynamics; only the DoD determines success.
Use only these terminal outcomes: . Never relabel a
non-success stop as done.
DONE | BLOCKED | CANCELLED | UNSAFE | IMPOSSIBLE | FAILED | EXHAUSTED | STOPPED | HANDOFF在第一次迭代前,绑定Focus Record、完善当前端点事实、变更和外部影响权限、配置文件、路由、状态类别、明确的Definition of Done(DoD)、验证器映射、有限活动预算、等待策略和主机能力。向reducer提供完整事实;遗漏的事实不会被视为已认证的假值。
每个循环仅保留一个主要焦点。将迭代定义为一次有限的、产生证据的状态转换——而非一次轮次、工具调用、编辑或勾选框。在每个边界处:
- 协调当前用户、工作区、验证器、子任务和外部事实。
- 使过期证据失效,并选择一个与强制缺口或决策相关不确定性绑定的就绪焦点。
- 在执行操作前记录当前焦点,仅对/
S2使用持久化协议。S3 - 通过至多一个主要专业领域执行一次授权的有限操作。
- 自动运行选定的验证器、受影响的回归测试和必需的集成检查。
- 记录结果、证据、成本、预算修订,以及恰好一个后续状态或类型化终端结果,然后再进行下一次选择。
DONEPASS在下一个循环前,需要一次授权操作,且该操作有合理、适配的机会填补强制缺口或解决实质性不确定性。切勿仅为消耗预算而重复未改变的因果策略。Convergence描述的是相对于目标的动态;只有DoD能决定成功与否。
仅使用以下终端结果:。切勿将非成功的终止重新标记为完成。
DONE | BLOCKED | CANCELLED | UNSAFE | IMPOSSIBLE | FAILED | EXHAUSTED | STOPPED | HANDOFFComposition
组合规则
Loop is a macro controller, not a primary specialty. Give one bounded iteration
at a time to at most one primary specialty. Nerd Execute owns implementation,
Nerd Surgery causal diagnosis and repair gates, and Nerd Patrol security
examination. Nerd Fast may optimize an iteration without weakening proof or
state. Nerd Memory is explicit-invocation-only admission middleware.
Loop是一个宏控制器,而非主要专业领域。每次将一个有限迭代交给至多一个主要专业领域处理。Nerd Execute负责实现,Nerd Surgery负责因果诊断和修复 gate,Nerd Patrol负责安全检查。Nerd Fast可在不弱化证明或状态的前提下优化迭代。Nerd Memory是仅明确调用的准入中间件。
Self-Refinement Loop
自我优化循环
When changing this skill, define the refinement DoD before editing, preserve
reducer fixtures, validate structure and behavior, and obtain fresh independent
review when it materially improves confidence. Finish only when mandatory
validation passes, no valid Blocker/High finding remains open, and two
successive fresh reviews produce no new valid Blocker/High finding. Reviewer
agreement alone is not proof, and repeated unchanged review is not progress.
Local readiness does not authorize installation, publication, deployment, or
another external release effect.
After changing this skill family, run .
python3 scripts/validate_skills.py修改此技能时,需在编辑前定义优化的DoD,保留reducer测试用例,验证结构和行为,并在能实质性提升可信度时获取新鲜的独立评审。仅当强制验证通过、无有效的Blocker/High级问题未解决,且连续两次新鲜评审未发现新的有效Blocker/High级问题时,才算完成。仅评审达成一致不足以证明成功,重复未改变的评审也不算进展。本地就绪状态不授权进行安装、发布、部署或其他外部发布操作。
修改此技能家族后,运行。
python3 scripts/validate_skills.py