knowledge

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Knowledge Skill

YOU MUST EXECUTE THIS WORKFLOW. Do not just describe it.

Find and retrieve knowledge from past work.

你必须执行此工作流程，而不只是描述它。

查找并检索过往工作中的知识。

Execution Steps

执行步骤

Given

/knowledge <query>

当收到

/knowledge <查询内容>

指令时：

Step 1: Search with ao CLI (if available)

步骤1：使用ao CLI进行搜索（如果可用）

bash

ao forge search "<query>" --limit 10 2>/dev/null

If results found, read the relevant files.

bash

ao forge search "<query>" --limit 10 2>/dev/null

如果找到结果，读取相关文件。

Step 2: Search .agents/ Directory

步骤2：搜索.agents/目录

bash

undefined

bash

undefined

Search learnings

搜索经验总结

grep -r "<query>" .agents/learnings/ 2>/dev/null | head -10

Search patterns

搜索模式

grep -r "<query>" .agents/patterns/ 2>/dev/null | head -10

Search research

搜索研究资料

grep -r "<query>" .agents/research/ 2>/dev/null | head -10

Search retros

搜索回顾文档

grep -r "<query>" .agents/retros/ 2>/dev/null | head -10

undefined

grep -r "<query>" .agents/retros/ 2>/dev/null | head -10

undefined

Step 3: Search Plans

步骤3：搜索计划文档

bash

undefined

bash

undefined

Local plans

本地计划

grep -r "<query>" .agents/plans/ 2>/dev/null | head -10

Global plans

全局计划

grep -r "<query>" ~/.claude/plans/ 2>/dev/null | head -10

undefined

grep -r "<query>" ~/.claude/plans/ 2>/dev/null | head -10

undefined

Step 4: Use Semantic Search (if MCP available)

步骤4：使用语义搜索（如果MCP可用）

Tool: mcp__smart-connections-work__lookup
Parameters:
  query: "<query>"
  limit: 10

Tool: mcp__smart-connections-work__lookup
Parameters:
  query: "<query>"
  limit: 10

Step 5: Read Relevant Files

步骤5：读取相关文件

For each match found, use the Read tool to get full content.

对于每个找到的匹配项，使用Read工具获取完整内容。

Step 6: Synthesize Results

步骤6：综合结果

Combine findings into a coherent response:

What do we know about this topic?
What learnings are relevant?
What patterns apply?
What past decisions were made?

将发现的内容整合成连贯的回复：

我们对该主题有哪些了解？
哪些经验总结相关？
适用哪些模式？
过往做出过哪些决策？

Step 7: Report to User

步骤7：向用户汇报

Present the knowledge found:

Summary of findings
Key learnings (with IDs)
Relevant patterns
Links to source files
Confidence level (how much we know)

呈现找到的知识：

发现内容摘要
关键经验总结（附ID）
相关模式
源文件链接
置信度（我们对该主题的了解程度）

Knowledge Locations

知识存储位置

Type	Location	Format
Learnings	`.agents/learnings/`	Markdown
Patterns	`.agents/patterns/`	Markdown
Research	`.agents/research/`	Markdown
Retros	`.agents/retros/`	Markdown
Plans	`.agents/plans/`	Markdown
Global Plans	`~/.claude/plans/`	Markdown

类型	位置	格式
经验总结	`.agents/learnings/`	Markdown
模式	`.agents/patterns/`	Markdown
研究资料	`.agents/research/`	Markdown
回顾文档	`.agents/retros/`	Markdown
计划	`.agents/plans/`	Markdown
全局计划	`~/.claude/plans/`	Markdown

Key Rules

核心规则

Search multiple locations - knowledge may be scattered
Use ao CLI first - semantic search is better
Fall back to grep - if ao not available
Read full files - don't just report matches
Synthesize - combine findings into useful answer

搜索多个位置 - 知识可能分散在各处
优先使用ao CLI - 语义搜索效果更佳
退而使用grep - 如果ao CLI不可用
读取完整文件 - 不要只汇报匹配的行
进行综合 - 将发现的内容整合为有用的答案

Example Queries

示例查询

bash

/knowledge authentication    # Find auth-related learnings
/knowledge "rate limiting"   # Find rate limit patterns
/knowledge kubernetes        # Find K8s knowledge
/knowledge "what do we know about caching"

bash

/knowledge authentication    # 查找与认证相关的经验总结
/knowledge "rate limiting"   # 查找限流模式
/knowledge kubernetes        # 查找K8s相关知识
/knowledge "what do we know about caching"

Examples

示例

Finding Past Learnings

查找过往经验总结

User says:

/knowledge "error handling patterns"

What happens:

Agent tries

ao forge search "error handling patterns"

, finds 3 matches

Agent searches
```
.agents/learnings/
```
with grep, finds 5 additional matches
Agent searches
```
.agents/patterns/
```
for related patterns, finds 2 matches
Agent reads all matched files using Read tool
Agent synthesizes findings into coherent response
Agent reports: "We have 5 learnings about error handling: L1 (always wrap errors), L3 (use typed errors), L12 (log before returning), L15 (context propagation), L22 (retry with backoff)"
Agent provides links to source files and confidence level: high (multiple confirmations)

Result: Complete knowledge synthesis with 5 specific learnings and 2 related patterns, all with source citations.

用户输入：

/knowledge "error handling patterns"

执行过程：

Agent尝试执行

ao forge search "error handling patterns"

，找到3个匹配项

Agent使用grep搜索
```
.agents/learnings/
```
，找到另外5个匹配项
Agent搜索
```
.agents/patterns/
```
查找相关模式，找到2个匹配项
Agent使用Read工具读取所有匹配的文件
Agent将发现的内容整合成连贯的回复
Agent汇报："我们有5条关于错误处理的经验总结：L1（始终包装错误）、L3（使用类型化错误）、L12（返回前先记录日志）、L15（上下文传播）、L22（退避重试）"
Agent提供源文件链接及置信度：高（多次确认）

结果： 完整的知识综合内容，包含5条具体经验总结和2个相关模式，均附有来源引用。

Querying Without ao CLI

无ao CLI时的查询

User says:

/knowledge "database migrations"

What happens:

Agent tries
```
ao forge search
```
, command not found
Agent falls back to grep search across
```
.agents/
```
directories
Agent finds 2 matches in learnings, 1 in research, 0 in patterns
Agent reads matched files
Agent synthesizes: "Limited knowledge found. L8 recommends using transaction-wrapped migrations. Research doc from 2026-01-20 analyzed migration tools."
Agent reports medium confidence (only 2 sources)

Result: Knowledge found despite missing ao CLI, with appropriate confidence level based on source count.

用户输入：

/knowledge "database migrations"

执行过程：

Agent尝试执行
```
ao forge search
```
，但命令未找到
Agent退而使用grep搜索
```
.agents/
```
目录
Agent在经验总结中找到2个匹配项，研究资料中找到1个，模式中无匹配
Agent读取匹配的文件
Agent综合内容："找到的知识有限。L8建议使用事务包装迁移。2026-01-20的研究文档分析了迁移工具。"
Agent汇报置信度：中等（仅2个来源）

结果： 尽管缺少ao CLI，仍找到了相关知识，并根据来源数量给出了相应的置信度。

Troubleshooting

故障排除

Problem	Cause	Solution
No results found	Query too specific or knowledge not yet captured	Broaden search terms. Try synonyms. Check if topic was covered in recent work but retro not yet run. Suggest running `/retro` to extract recent learnings.
Too many results (overwhelming)	Very broad query term	Narrow query with more specific terms. Filter by date: search only recent learnings. Use semantic search (ao CLI) for better ranking if available.
Results lack context	Grep matches found but files don't address query	Read full files, not just matching lines. Synthesize from surrounding context. May need to trace back to original research with `/trace` .
Confidence level unclear	Mixed or contradictory sources	Report conflicting information explicitly. Note which sources agree/disagree. Suggest running `/research` to investigate further if critical.

问题	原因	解决方案
未找到结果	查询过于具体或知识尚未被记录	拓宽搜索词，尝试同义词。检查该主题是否在近期工作中涉及但尚未完成回顾。建议运行 `/retro` 提取近期经验总结。
结果过多（难以处理）	查询词过于宽泛	使用更具体的词缩小查询范围。按日期筛选：仅搜索近期经验总结。如果可用，使用语义搜索（ao CLI）获得更好的排序结果。
结果缺乏上下文	找到grep匹配项但文件未回应查询	读取完整文件，而非仅匹配行。结合上下文进行综合。可能需要使用 `/trace` 追溯到原始研究资料。
置信度不明确	来源混杂或存在矛盾	明确汇报相互矛盾的信息。注明哪些来源达成一致/存在分歧。如果问题关键，建议运行 `/research` 进一步调查。