Skip to content

Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.

优化yaml;改为英文;增加name规则 (#2677)

DuTaoc4944d28d74a
diff --git a/openviking/prompts/templates/skill_extract/session_skills.yaml b/openviking/prompts/templates/skill_extract/session_skills.yaml
index 71b4832a8..3620fc828 100644
--- a/openviking/prompts/templates/skill_extract/session_skills.yaml
+++ b/openviking/prompts/templates/skill_extract/session_skills.yaml
@@ -1,105 +1,105 @@
memory_type: session_skills
description: |
- # 任务
- 你的任务是基于对话记录,记录下用户主动要求记录的会话技能,步骤规范如下:
- 一、提取内容的识别规则
- (一)显式指令识别规则
- 1. 触发词识别:识别 “以后”“今后”“每次” 等时间类触发词,判定为长期记忆指令。
- 2. 规范性表达识别:识别 “按照...”“遵循...” 等表述,提取其中的规范要求作为记忆规则。
- 3. 直接请求识别:识别 “记住”“别忘了” 等直接记忆请求,明确需记忆的核心内容。
- (二)隐式指令识别规则
- 1. 修正意图识别:从用户对之前回应的修正内容中,提取用户偏好并判定为可记忆规则。
- 2. 重复模式识别:当用户在不同时间(如不同日期)多次重复相同要求(≥3 次),自动识别为隐式记忆指令。
- (三)可记忆性判断规则
- 1. 时间 / 范围限定判断:含 “今天”“这次” 等一次性时间限定的为临时要求,不记忆;含 “所有”“每次” 等无时间限制或范围宽泛表述的为长期规范,需记忆。
- 2. 内容类型判断:具体题目答案、单次任务结果等具体内容不记忆;选择题标注答案、逐选项分析等通用规则需记忆。
- 3. 适用范围判断:表述模糊(如 “要详细一点”)、适用范围不明确的,需向用户确认后再判断是否记忆。
- (四)增量更新识别规则
- 1. 新增意图识别:通过 “对了” 等关键词识别用户追加规则的意图,检查与现有规则的兼容性后整合。
- 2. 修正意图识别:通过 “不对”“改一下” 等关键词识别规则修正意图,精准定位被修改的原规则。
- 3. 删除意图识别:通过 “不需要”“取消” 等关键词识别规则删除意图,同时检查被删除规则是否有依赖规则。
- (五)冲突识别规则
- 1. 直接矛盾识别:新规则与现有规则存在明显对立表述(如 “简洁” vs “详细”),判定为直接冲突。
- 2. 逻辑冲突识别:结合场景计算后出现矛盾(如单选项 100 字 + 要求与总解析 300 字限制),判定为逻辑冲突。
- 3. 范围冲突识别:全局规则与具体任务规则存在格式、要求等冲突(如全局 Markdown 格式与 PPT 制作的 PPTX 格式),判定为范围冲突。
- (六)任务匹配识别规则
- 1. 任务类型识别:明确当前任务的核心类型(如题目分析、周报撰写)和所属领域(如医学、财务),匹配对应记忆规则。
- 2. 相似度分级判断:任务类型与记忆规则完全一致(如医学题分析匹配医学题规则)为高相似度(95%+),直接应用;任务类型相似(如月度总结匹配周报规则)为中等相似度(60-80%),需用户确认后应用;任务类型无关(如写邮件匹配医学题规则)为低相似度(<30%),不应用。
- 二、提取内容的格式规范
- (一)指令提取格式规范
- 1. 约束条件格式:需明确 “约束类型 + 具体要求”,如 “语气约束:正式、专业”“表达约束:简洁,避免冗余”。
- 2. 格式要求格式:按 “格式类型 + 具体规范” 提取,如 “Markdown 格式:表格用三线表、代码需高亮”。
- 3. 处理步骤格式:按 “步骤序号 + 步骤内容” 提取,保持原顺序,标注 “处理流程类”,如 “步骤 1:识别题型;步骤 2:列知识点;步骤 3:逐步解析;步骤 4:给类似题”。
- 4. 质量标准格式:按 “标准类型 + 具体要求” 提取,标注 “质量标准类”,如 “清晰度标准:易懂、学生可理解;准确性标准:无错误”。
- (二)指令规范化格式规范
- 1. 口语化转标准化:将 “要那种正式的感觉” 转化为 “语气正式、专业”,“别太啰嗦” 转化为 “表达简洁,避免冗余”,去除口语化表述。
- 2. 模糊表述量化:对 “详细”“几个” 等模糊词,通过询问用户确认后量化,如 “详细→解析长度:80-150 字”“几个→2-3 个”。
- 3. 隐含信息显式化:识别 “上次” 等模糊指代,查找历史记录补全具体内容,明确完整规则,不保留模糊表述。
- 4. 冗余信息处理:合并重复表述,如 “每个选项都要分析,不要把选项混在一起,要一个一个分析” 合并为 “逐选项单独分析”,保留核心语义。
- (三)指令分类格式规范
- 1. 功能分类格式:按 “分类名称:规则内容” 分类,如 “输入处理类:2 条;核心逻辑类:4 条;输出格式类:3 条;约束禁忌类:1 条”。
- 2. 条件分支格式:按 “条件 + 执行动作” 表述,如 “条件 1:选择题→标注答案;条件 2:简答题→给要点”。
- (四)适用范围标注格式规范
- 1. 任务范围格式:明确 “规则内容 + 适用范围 + 排除范围”,如 “规则:标注知识点来源;适用范围:医学题;排除范围:编程题”。
- 2. 条件范围格式:按 “条件 + 规则内容” 表述,如 “条件:小学生作文→规则:鼓励性语言;条件:高中作文→规则:客观专业”。
- 3. 作用域格式:标注 “全局指令 / 具体意图指令”,明确层级关系,如 “全局指令:所有文档都用 Markdown;具体意图指令:周报要包含三部分”。
- (五)冲突解决格式规范
- 1. 条件分支解决格式:按 “场景 + 规则内容” 明确分支,如 “正确选项:简洁(30-50 字);错误选项:详细解释(80-120 字)”。
- 2. 优先级标注格式:明确 “优先级顺序:当前指令> 任务规则 > 全局规则”,标注执行方式和记忆保留策略。
- 3. 范围细化格式:细化规则条件和要求,确保无逻辑冲突,如 “关键选项(正确选项 + 易错选项):80 字;其他选项:40 字”。
- 三、例子
- (一)显式指令识别例子
- - 输入:“以后分析医学题时,都要标注知识点来源”
- - 识别结果:显式指令(触发词 “以后”),提取规则 “标注知识点来源”,判定可记忆,适用范围为医学题分析。
- (二)隐式指令识别例子
- - 输入:
- - 轮次 1 用户:“帮我分析这道题”
- - 轮次 1 Agent:[给出未解释错误选项的分析]
- - 轮次 2 用户:“不对,错误选项要解释为什么错”
- - 识别结果:隐式指令(用户修正),判定可记忆,询问用户 “您希望以后分析题目时,都这样处理错误选项吗?”
- (三)可记忆性判断例子
- - 例子 1(一次性要求):
- - 输入:“今天的报告要加上财务数据”
- - 判断结果:一次性要求(时间限定 “今天”),不记忆。
- - 例子 2(长期规范):
- - 输入:“所有报告都要包含财务数据”
- - 判断结果:长期规范(范围 “所有”),记忆。
- - 例子 3(模糊表述):
- - 输入:“要详细一点”
- - 判断结果:适用范围不明,询问用户 “您是指所有任务都要详细,还是只针对某类任务?”
- (四)指令提取与规范化例子
- - 输入:“要那种正式的感觉,别太啰嗦,重点突出”
- - 提取结果:
- - 规则 1:语气约束:正式、专业
- - 规则 2:表达约束:简洁,避免冗余
- - 规则 3:格式要求:重点内容前置或加粗强调
- - 规范化结果:去除口语化表述,明确约束类型和具体要求。
- (五)增量更新例子
- - 例子 1(新增规则):
- - 现有记忆:规则 1 “逐选项分析”,规则 2 “标注来源”
- - 输入:“对了,错误选项要指出常见误区”
- - 处理结果:识别追加意图,新增规则 3 “错误选项标注常见误区”,与现有规则无冲突,整合为规则 1-2-3。
- - 例子 2(修正规则):
- - 现有:“解析长度:50 字以内”
- - 输入:“不对,应该是 50-100 字”
- - 处理结果:识别修正意图,定位 “解析长度” 规则,替换为 “50-100 字”,验证与其他规则一致性。
- (六)冲突处理例子
- - 输入:
- - 现有:“解析要简洁”
- - 新增:“错误选项要详细解释”
- - 冲突检测:直接冲突(“简洁” vs “详细”)
- - 解决询问:“您之前要求简洁,现在又要详细解释错误选项,您希望:A. 只有错误选项详细,其他部分简洁;B. 根据情况灵活处理;C. 修改之前的 ' 简洁 ' 要求”
- - 最终规则(用户选 A):“正确选项:简洁(30-50 字);错误选项:详细解释(80-120 字)”
- (七)任务匹配与记忆应用例子
- - 例子 1(高相似度匹配):
- - 记忆:医学题分析规则(标注来源、逐选项分析等)
- - 输入:“帮我分析这道药理学题目”
- - 匹配结果:相似度 95%+,直接应用医学题分析规则。
- - 例子 2(跨会话一致性应用):
- - Day1:用户设定规则 “解析 50-100 字,标注来源”
- - Day3:输入 “分析这道新题”
- - Day7:输入 “再分析一道题”
- - 应用结果:Day3 和 Day7 自动应用 Day1 的规则,三次输出格式、细节保持一致。
+ # Task
+ Your task is to record session skills that the user explicitly asks to remember based on the conversation history. Follow these rules:
+ 1. Recognition rules for extracted content
+ 1.1 Explicit instruction recognition
+ 1. Trigger word recognition: Recognize time-related trigger words such as "from now on", "in the future", and "every time", and treat them as long-term memory instructions.
+ 2. Normative expression recognition: Recognize expressions such as "follow..." and "according to...", and extract the normative requirements as memory rules.
+ 3. Direct request recognition: Recognize direct memory requests such as "remember" and "don't forget", and identify the core content that needs to be remembered.
+ 1.2 Implicit instruction recognition
+ 1. Correction intent recognition: Extract user preferences from corrections to previous responses and determine whether they are memorable rules.
+ 2. Repeated pattern recognition: When the user repeats the same requirement multiple times at different times, such as on different dates, and the count is at least 3, automatically recognize it as an implicit memory instruction.
+ 1.3 Memorability judgment
+ 1. Time and scope limitation judgment: Requirements with one-time time limits such as "today" or "this time" are temporary and should not be remembered. Requirements with no time limit or broad scope such as "all" or "every time" are long-term norms and should be remembered.
+ 2. Content type judgment: Specific content such as answers to individual questions or one-off task results should not be remembered. General rules such as marking answers for multiple-choice questions or analyzing each option should be remembered.
+ 3. Applicability judgment: For vague expressions such as "make it more detailed" or unclear applicability scope, ask the user for confirmation before deciding whether to remember the rule.
+ 1.4 Incremental update recognition
+ 1. Addition intent recognition: Use keywords such as "by the way" to identify the user's intent to add rules. Check compatibility with existing rules before merging.
+ 2. Revision intent recognition: Use keywords such as "that's not right" or "change it" to identify rule revision intent and precisely locate the original rule being modified.
+ 3. Deletion intent recognition: Use keywords such as "no need" or "cancel" to identify rule deletion intent, and also check whether the deleted rule has dependent rules.
+ 1.5 Conflict recognition
+ 1. Direct contradiction recognition: If a new rule and an existing rule contain clearly opposing requirements, such as "concise" vs "detailed", treat it as a direct conflict.
+ 2. Logical conflict recognition: If requirements become contradictory after considering the scenario, such as 100 words per option plus a 300-word limit for the overall analysis, treat it as a logical conflict.
+ 3. Scope conflict recognition: If a global rule conflicts with a specific task rule in format or requirements, such as a global Markdown format requirement vs a PPTX format requirement for slide creation, treat it as a scope conflict.
+ 1.6 Task matching recognition
+ 1. Task type recognition: Identify the core type of the current task, such as question analysis or weekly report writing, and its domain, such as medicine or finance, then match the corresponding memory rule.
+ 2. Similarity tier judgment: If the task type exactly matches the memory rule, such as medical question analysis matching medical question rules, treat it as high similarity, 95% or above, and apply it directly. If the task types are similar, such as monthly summaries matching weekly report rules, treat it as medium similarity, 60-80%, and ask the user for confirmation before applying it. If the task types are unrelated, such as writing an email matching medical question rules, treat it as low similarity, below 30%, and do not apply it.
+ 2. Formatting rules for extracted content
+ 2.1 Instruction extraction format
+ 1. Constraint format: Clearly state "constraint type + specific requirement", such as "tone constraint: formal and professional" or "expression constraint: concise, avoid redundancy".
+ 2. Format requirement format: Extract as "format type + specific specification", such as "Markdown format: use three-line tables and highlight code".
+ 3. Processing step format: Extract as "step number + step content", preserve the original order, and mark it as a process flow type. For example: "Step 1: identify the question type; Step 2: list knowledge points; Step 3: analyze step by step; Step 4: provide similar questions".
+ 4. Quality standard format: Extract as "standard type + specific requirement", mark it as a quality standard type. For example: "clarity standard: easy to understand and suitable for students; accuracy standard: no errors".
+ 2.2 Instruction normalization format
+ 1. Convert colloquial phrasing to standard phrasing: Convert "make it feel formal" to "formal and professional tone", and "don't be too wordy" to "concise expression, avoid redundancy". Remove colloquial expressions.
+ 2. Quantify vague expressions: For vague words such as "detailed" or "several", ask the user for confirmation before quantifying them, such as "detailed -> analysis length: 80-150 words" or "several -> 2-3".
+ 3. Make implicit information explicit: Recognize vague references such as "last time", find the specific content from history, complete the rule, and do not keep vague expressions.
+ 4. Handle redundancy: Merge repeated expressions, such as converting "analyze every option, don't mix the options together, analyze them one by one" into "analyze each option separately", while preserving the core meaning.
+ 2.3 Instruction classification format
+ 1. Functional classification format: Classify as "category name: rule content", such as "input handling: 2 rules; core logic: 4 rules; output format: 3 rules; constraints and prohibitions: 1 rule".
+ 2. Conditional branch format: Express as "condition + action", such as "Condition 1: multiple-choice question -> mark the answer; Condition 2: short-answer question -> provide key points".
+ 2.4 Applicability scope annotation format
+ 1. Task scope format: Clearly state "rule content + applicable scope + excluded scope", such as "rule: mark knowledge source; applicable scope: medical questions; excluded scope: programming questions".
+ 2. Conditional scope format: Express as "condition + rule content", such as "condition: elementary school essay -> rule: encouraging language; condition: high school essay -> rule: objective and professional".
+ 3. Scope level format: Mark "global instruction / specific intent instruction" and clearly state the hierarchy, such as "global instruction: all documents use Markdown; specific intent instruction: weekly reports must include three parts".
+ 2.5 Conflict resolution format
+ 1. Conditional branch resolution format: Clearly state branches as "scenario + rule content", such as "correct option: concise, 30-50 words; incorrect option: detailed explanation, 80-120 words".
+ 2. Priority annotation format: Clearly state "priority order: current instruction > task rule > global rule", and annotate the execution method and memory retention strategy.
+ 3. Scope refinement format: Refine rule conditions and requirements to avoid logical conflicts, such as "key options, including correct and easily confused options: 80 words; other options: 40 words".
+ 3. Examples
+ 3.1 Explicit instruction recognition example
+ - Input: "From now on, when analyzing medical questions, always mark the knowledge source."
+ - Recognition result: Explicit instruction, triggered by "from now on". Extract the rule "mark the knowledge source", determine that it is memorable, and set the applicable scope to medical question analysis.
+ 3.2 Implicit instruction recognition example
+ - Input:
+ - Round 1 user: "Help me analyze this question."
+ - Round 1 Agent: [provides an analysis that does not explain why the incorrect options are wrong]
+ - Round 2 user: "That's not right. For incorrect options, explain why they are wrong."
+ - Recognition result: Implicit instruction, from user correction. Determine that it is memorable and ask the user: "Would you like me to handle incorrect options this way whenever analyzing questions in the future?"
+ 3.3 Memorability judgment examples
+ - Example 1, one-time requirement:
+ - Input: "Add financial data to today's report."
+ - Judgment result: One-time requirement with the time limit "today"; do not remember it.
+ - Example 2, long-term norm:
+ - Input: "All reports must include financial data."
+ - Judgment result: Long-term norm with the scope "all"; remember it.
+ - Example 3, vague expression:
+ - Input: "Make it more detailed."
+ - Judgment result: Applicability scope is unclear. Ask the user: "Do you mean all tasks should be more detailed, or only a certain type of task?"
+ 3.4 Instruction extraction and normalization example
+ - Input: "Make it feel formal, don't be too wordy, and highlight the key points."
+ - Extraction result:
+ - Rule 1: Tone constraint: formal and professional
+ - Rule 2: Expression constraint: concise, avoid redundancy
+ - Rule 3: Format requirement: put key content first or emphasize it in bold
+ - Normalization result: Remove colloquial expressions and clearly state the constraint types and specific requirements.
+ 3.5 Incremental update examples
+ - Example 1, adding a new rule:
+ - Existing memory: Rule 1 "analyze each option"; Rule 2 "mark source"
+ - Input: "By the way, incorrect options should mention common misconceptions."
+ - Processing result: Identify the addition intent, add Rule 3 "mark common misconceptions for incorrect options", confirm there is no conflict with existing rules, and merge as Rules 1-2-3.
+ - Example 2, revising a rule:
+ - Existing: "analysis length: within 50 words"
+ - Input: "That's not right, it should be 50-100 words."
+ - Processing result: Identify revision intent, locate the "analysis length" rule, replace it with "50-100 words", and verify consistency with other rules.
+ 3.6 Conflict handling example
+ - Input:
+ - Existing: "The analysis should be concise."
+ - New: "Incorrect options should be explained in detail."
+ - Conflict detection: Direct conflict, "concise" vs "detailed".
+ - Resolution question: "You previously asked for concise analysis, and now you want detailed explanations for incorrect options. Which do you prefer: A. Only incorrect options are detailed while other parts stay concise; B. handle flexibly based on the situation; C. modify the previous 'concise' requirement?"
+ - Final rule, if the user selects A: "correct option: concise, 30-50 words; incorrect options: detailed explanation, 80-120 words"
+ 3.7 Task matching and memory application examples
+ - Example 1, high-similarity match:
+ - Memory: Medical question analysis rules, such as marking sources and analyzing each option.
+ - Input: "Help me analyze this pharmacology question."
+ - Match result: Similarity 95% or above; directly apply the medical question analysis rules.
+ - Example 2, cross-session consistency:
+ - Day 1: User sets the rule "analysis should be 50-100 words and mark sources"
+ - Day 3: Input "Analyze this new question"
+ - Day 7: Input "Analyze another question"
+ - Application result: Day 3 and Day 7 automatically apply the Day 1 rule, and all three outputs keep the same format and level of detail.
directory: "viking://user/{{ user_space }}/skills"
filename_template: "{{ skill_name }}/SKILL.md"
enabled: true
@@ -107,79 +107,91 @@ fields:
- name: skill_name
type: string
description: |
- # 内容
- 可应用上该可Skill的场景描述的中文名
- # 注意事项
- (1)要尽可能的精简,避免使用复杂的描述词汇。不超过12个字符;
- (2)不得为空。
- # 例子
- ## 好例子
- - 订单处理流程
- ## 坏例子
- - 订单处理的详细流程,尤其是退款订单
+ # Content
+ The stable name of the Skill.
+ # Notes
+ (1) It must be a non-empty string.
+ (2) It must not exceed 64 characters.
+ (3) It may contain only ASCII letters, digits, underscores `_`, and hyphens `-`.
+ (4) It must not contain Chinese characters, spaces, slashes, dots, colons, or other punctuation.
+ (5) Prefer lowercase English words connected with hyphens. Keep it short while expressing the main purpose.
+ # Examples
+ ## Good examples
+ - order-refund-flow
+ - medical-question-analysis
+ - report_format_rules
+ ## Bad examples
+ - order-refund-flow!
+ - order refund flow
+ - order/refund/flow
+ - order.refund.flow
+ - this-skill-name-is-too-long-and-contains-too-many-unnecessary-details-for-storage
merge_op: immutable
- name: description
type: string
description: |
- # 内容 简要介绍该技能功能或应用场景
- # 注意事项 (1)输出必须是通用的,不涉及具体要求;(2)不超过30字符;
- # 例子
- ## 好例子
- - 分析用户常识问题时的详细步骤,尤其是地理问题 胃痛症状如何进行分析和处理,包括检查症状、咨询医生、建议治疗等步骤。
- ## 坏例子
- - 用户反馈快递未收到但系统显示已完成,要求解决问题(拿到货或退款),并询问订单12346配送情况
+ # Content
+ Briefly describe the skill's function or application scenario.
+ # Notes
+ (1) The output must be general and must not include specific task requirements.
+ (2) It must not exceed 30 characters.
+ # Examples
+ ## Good examples
+ - Detailed steps for analyzing common-sense questions, especially geography questions. How to analyze and handle stomach pain symptoms, including checking symptoms, consulting a doctor, and suggesting treatment.
+ ## Bad examples
+ - The user reports that the package was not received even though the system shows delivery completed, asks to resolve the issue by getting the item or a refund, and asks about the delivery status of order 12346.
merge_op: replace
- name: content
type: string
description: |
- # 内容
- 该Skill的完整信息,必须与从用户和模型的对话记录中提取的信息保持一致。
+ # Content
+ The complete information for this Skill. It must remain consistent with the information extracted from the conversation history between the user and the model.
- # 输出格式
- 1.必须以连续的Markdown字符串格式进行输出;
- 2.可参考给出的格式说明,部分章节可按情况调整
+ # Output format
+ 1. The output must be a continuous Markdown string.
+ 2. You may refer to the format description below and adjust some sections as appropriate.
- # 详细格式和说明:参考如下<md>标签里的示例Markdown内容
+ # Detailed format and instructions: refer to the example Markdown content inside the <md> tag below.
<md>
- [简要介绍该技能及其重要性]
- ## 核心规范
- - [核心主题 1]
- - [核心主题 2]
- ## 使用该Skill的场景
- - [场景 1]
- - [场景 2]
- ## 实现指南
- ### 先决条件
- - [所需的知识或者环境、配置]
- ### 技术和方法
- [详细说明如何运用此技能]
- ### 应用场景示例
- [来自原始材料的具体实例]
- ## 核心要点
- - [要点 1]
- - [要点 2]
+ [Briefly introduce the skill and its importance]
+ ## Core Rules
+ - [Core topic 1]
+ - [Core topic 2]
+ ## Scenarios for Using This Skill
+ - [Scenario 1]
+ - [Scenario 2]
+ ## Implementation Guide
+ ### Prerequisites
+ - [Required knowledge, environment, or configuration]
+ ### Techniques and Methods
+ [Explain in detail how to apply this skill]
+ ### Application Scenario Examples
+ [Concrete examples from the source material]
+ ## Key Points
+ - [Key point 1]
+ - [Key point 2]
</md>
- # 正确输出参考如下<example>标签里的内容
+ # Correct output reference: see the content inside the <example> tag below.
<example>
- 本技能用于规范物理电路计算题的解答流程,确保解答的准确性和规范性。
- ## 核心规范
- - 解答物理电路计算题需按标准格式进行
- - 标准格式包含电路分析、列公式、计算和输出结果四个步骤
- ## 使用该Skill的场景
- - 解答物理电路计算题时
- ## 实现指南
- ### 先决条件
- - 掌握物理电路的基本概念和公式
- ### 技术和方法
- 1. 【电路分析】:用文字描述连接方式(串联/并联);
- 2. 【公式】:列出使用的原始公式;
- ### 应用场景示例
- 当遇到物理电路计算题时,按照上述四个步骤进行解答。
- ## 核心要点
- - 严格按照四个步骤解答
- - 计算过程不带单位,结果带单位
+ This skill standardizes the solution process for physics circuit calculation problems to ensure accuracy and consistency.
+ ## Core Rules
+ - Physics circuit calculation problems must be solved using the standard format
+ - The standard format contains four steps: circuit analysis, formula listing, calculation, and final answer output
+ ## Scenarios for Using This Skill
+ - When solving physics circuit calculation problems
+ ## Implementation Guide
+ ### Prerequisites
+ - Master the basic concepts and formulas of physics circuits
+ ### Techniques and Methods
+ 1. [Circuit analysis]: Describe the connection method in words, such as series or parallel.
+ 2. [Formula]: List the original formulas used.
+ ### Application Scenario Examples
+ When encountering a physics circuit calculation problem, solve it according to the four steps above.
+ ## Key Points
+ - Strictly follow the four-step solution process
+ - Do not include units in the calculation process; include units in the final answer
</example>
merge_op: patch