transcript-search
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseTranscript Search
会议记录搜索
Overview
概述
Search and retrieve meeting transcripts from Fireflies.ai and Google Drive with full text and speaker attribution.
从Fireflies.ai和Google Drive中搜索并检索会议记录,支持全文检索和发言者归属标记。
Quick Decision Tree
快速决策树
Where are your transcripts?
│
├── Fireflies.ai (API)
│ └── references/fireflies.md
│ └── Script: scripts/fireflies_transcript_search.py
│
└── Google Drive (uploaded files)
└── references/drive-transcripts.md
└── Script: scripts/gdrive_transcript_search.pyWhere are your transcripts?
│
├── Fireflies.ai (API)
│ └── references/fireflies.md
│ └── Script: scripts/fireflies_transcript_search.py
│
└── Google Drive (uploaded files)
└── references/drive-transcripts.md
└── Script: scripts/gdrive_transcript_search.pyEnvironment Setup
环境配置
bash
undefinedbash
undefinedFor Fireflies
For Fireflies
FIREFLIES_API_KEY=your_api_key
FIREFLIES_API_KEY=your_api_key
For Google Drive
For Google Drive
OAuth credentials (mycreds.txt) - see google-workspace skill
OAuth credentials (mycreds.txt) - see google-workspace skill
Get Fireflies API key: https://app.fireflies.ai/integrations (Custom Integrations)
获取Fireflies API密钥:https://app.fireflies.ai/integrations (Custom Integrations)Common Usage
常见用法
Search Fireflies by Company
按公司名称搜索Fireflies记录
bash
python scripts/fireflies_transcript_search.py "Microsoft" --days-back 30bash
python scripts/fireflies_transcript_search.py "Microsoft" --days-back 30Get Full Transcript
获取完整会议记录
bash
python scripts/fireflies_transcript_search.py --id 01KCM2G0YX1GMPWYQ8GPAABBCK --contentbash
python scripts/fireflies_transcript_search.py --id 01KCM2G0YX1GMPWYQ8GPAABBCK --contentSave Formatted Transcript
保存格式化会议记录
bash
python scripts/fireflies_transcript_search.py --id abc123 --content --save transcript.mdbash
python scripts/fireflies_transcript_search.py --id abc123 --content --save transcript.mdSearch Drive Transcripts
搜索Drive中的会议记录
bash
python scripts/gdrive_transcript_search.py "Acme Corp" --days 30bash
python scripts/gdrive_transcript_search.py "Acme Corp" --days 30Output Formats
输出格式
Search Results
搜索结果
- Transcript ID, title, date, duration
- AI-generated summary, keywords, action items
- Speaker list with names
- 记录ID、标题、日期、时长
- AI生成的摘要、关键词、行动项
- 发言者姓名列表
Full Transcript
完整会议记录
- Complete text with speaker labels
- Timestamps for each sentence
- Formatted markdown output
- 带有发言者标签的完整文本
- 每句话的时间戳
- 格式化的Markdown输出
Cost
成本
| Service | Cost |
|---|---|
| Fireflies API | Free (with subscription) |
| Google Drive | Free |
| 服务 | 费用 |
|---|---|
| Fireflies API | 免费(需订阅基础服务) |
| Google Drive | 免费 |
Security Notes
安全注意事项
Credential Handling
凭证处理
- Store in
FIREFLIES_API_KEYfile (never commit to git).env - Google OAuth credentials for Drive (see google-workspace skill)
- Regenerate Fireflies API key via Custom Integrations if compromised
- Never log or print API keys in script output
- 将存储在
FIREFLIES_API_KEY文件中(绝不要提交到git).env - Google Drive的OAuth凭证(mycreds.txt)- 参考google-workspace技能
- 若凭证泄露,通过Custom Integrations重新生成Fireflies API密钥
- 绝不要在脚本输出中记录或打印API密钥
Data Privacy
数据隐私
- Transcripts contain verbatim meeting conversations
- Speaker names and voices are identifiable
- Business discussions may include confidential information
- Action items and summaries capture sensitive decisions
- Avoid sharing full transcripts without authorization
- 会议记录包含会议对话的原文
- 发言者姓名和声音可被识别
- 业务讨论可能包含机密信息
- 行动项和摘要会捕获敏感决策
- 未经授权请勿分享完整会议记录
Access Scopes
访问权限范围
- Fireflies API: Full access to transcripts user has access to
- Google Drive: Access to uploaded transcript files
- Transcripts inherit meeting participants' access permissions
- Fireflies API:可访问用户有权限查看的所有记录
- Google Drive:可访问已上传的会议记录文件
- 会议记录继承会议参与者的访问权限
Compliance Considerations
合规性考量
- Recording Consent: Ensure all meeting participants consent to recording
- Speaker Privacy: Transcripts identify speakers by name
- Confidential Meetings: Some meetings should not be transcribed
- GDPR: Meeting recordings containing EU participants require consent
- Data Retention: Follow policies for transcript retention/deletion
- Legal Holds: Transcripts may be subject to legal discovery
- Client Confidentiality: Client meeting content is sensitive
- Internal Use Only: Mark transcripts as confidential where appropriate
- 录音同意:确保所有会议参与者同意被录音
- 发言者隐私:会议记录会显示发言者姓名
- 机密会议:部分会议不应被转录
- GDPR:包含欧盟参与者的会议录音需获得同意
- 数据保留:遵循会议记录保留/删除政策
- 法律保全:会议记录可能需配合法律调查
- 客户机密性:客户会议内容属于敏感信息
- 仅限内部使用:必要时标记会议记录为机密
Troubleshooting
故障排除
Common Issues
常见问题
Issue: Transcript not found
问题:找不到会议记录
Symptoms: "Transcript not found" error with known meeting
Cause: Invalid transcript ID, no access, or not yet processed
Solution:
- Verify transcript ID from Fireflies dashboard
- Check if recording is still processing (wait and retry)
- Ensure API key has access to the transcript
- Confirm meeting was actually recorded and transcribed
症状:已知存在的会议却显示“Transcript not found”错误
原因:记录ID无效、无访问权限或尚未处理完成
解决方案:
- 从Fireflies控制面板验证记录ID
- 检查录音是否仍在处理中(等待后重试)
- 确保API密钥有权限访问该记录
- 确认会议确实被录音并转录
Issue: API unauthorized
问题:API未授权
Symptoms: 401 error or "invalid API key"
Cause: API key expired, invalid, or not set
Solution:
- Regenerate API key at https://app.fireflies.ai/integrations
- Verify is set correctly in
FIREFLIES_API_KEY.env - Check for leading/trailing whitespace in key
- Ensure Custom Integration is enabled in Fireflies
症状:401错误或“invalid API key”提示
原因:API密钥过期、无效或未正确设置
解决方案:
- 在https://app.fireflies.ai/integrations重新生成API密钥
- 验证文件中的
.env设置正确FIREFLIES_API_KEY - 检查密钥是否存在首尾空格
- 确保Fireflies中的Custom Integration已启用
Issue: Empty search results
问题:搜索结果为空
Symptoms: Search returns no transcripts despite existing meetings
Cause: Search query too narrow, date range issue, or access restrictions
Solution:
- Broaden search query (fewer keywords)
- Expand date range with parameter
--days-back - Check if transcripts are shared with your account
- Verify search is using correct field (title, participants, content)
症状:明明存在会议,但搜索无结果返回
原因:搜索条件过窄、日期范围问题或访问限制
解决方案:
- 放宽搜索条件(减少关键词)
- 使用参数扩大日期范围
--days-back - 检查记录是否已共享到你的账户
- 验证搜索是否使用了正确的字段(标题、参与者、内容)
Issue: Missing transcript content
问题:缺少会议记录内容
Symptoms: Metadata returned but full content is empty
Cause: Transcript still processing or content access restricted
Solution:
- Wait for transcript processing to complete (check Fireflies dashboard)
- Use flag explicitly to request full transcript
--content - Verify subscription tier includes API content access
- Check if transcript has processing errors
症状:返回元数据但完整内容为空
原因:记录仍在处理中或内容访问受限
解决方案:
- 等待记录处理完成(查看Fireflies控制面板)
- 显式使用参数请求完整记录
--content - 验证订阅套餐是否包含API内容访问权限
- 检查记录是否存在处理错误
Issue: Speaker attribution missing
问题:缺少发言者归属标记
Symptoms: Transcript text present but speakers not identified
Cause: Low audio quality or speakers not enrolled
Solution:
- Speaker identification depends on audio quality
- Enroll frequent speakers in Fireflies for better recognition
- This is a Fireflies processing issue, not API issue
- Re-upload recording if possible with better audio
症状:会议记录文本存在但未识别出发言者
原因:音频质量差或发言者未注册
解决方案:
- 发言者识别依赖音频质量
- 在Fireflies中注册频繁发言者以提升识别准确率
- 这是Fireflies的处理问题,而非API问题
- 若可能,重新上传音频质量更好的录音
Issue: Google Drive transcript search fails
问题:Google Drive会议记录搜索失败
Symptoms: Can't find transcripts stored in Drive
Cause: OAuth issue, folder structure, or file format
Solution:
- Verify Google OAuth is working (see google-workspace skill)
- Check transcripts are in searchable folder
- Ensure transcripts are in readable format (txt, md, docx)
- Search by exact filename if full-text search fails
症状:无法找到存储在Drive中的记录
原因:OAuth问题、文件夹结构或文件格式
解决方案:
- 验证Google OAuth是否正常工作(参考google-workspace技能)
- 检查记录是否在可搜索的文件夹中
- 确保记录为可读取格式(txt、md、docx)
- 若全文搜索失败,尝试按精确文件名搜索
Resources
参考资源
- references/fireflies.md - Fireflies.ai API guide
- references/drive-transcripts.md - Drive transcript search
- references/fireflies.md - Fireflies.ai API指南
- references/drive-transcripts.md - Drive会议记录搜索指南
Integration Patterns
集成模式
Transcript to Summary
记录转摘要
Skills: transcript-search → content-generation
Use case: Create meeting summaries and action item docs
Flow:
- Retrieve full transcript from Fireflies
- Extract key discussion points and decisions
- Generate formatted summary document via content-generation
技能: transcript-search → content-generation
使用场景: 创建会议摘要和行动项文档
流程:
- 从Fireflies获取完整会议记录
- 提取关键讨论点和决策
- 通过content-generation生成格式化的摘要文档
Transcript to Voice Agent
记录转语音Agent
Skills: transcript-search → voice-agents
Use case: Build context-aware voice agents
Flow:
- Search for past meetings with client
- Extract relationship history and previous discussions
- Include context in voice agent prompt for personalized calls
技能: transcript-search → voice-agents
使用场景: 构建上下文感知的语音Agent
流程:
- 搜索与客户的过往会议记录
- 提取合作历史和之前的讨论内容
- 将上下文纳入语音Agent的提示词,实现个性化通话
Transcript to CRM
记录转CRM
Skills: transcript-search → attio-crm
Use case: Add meeting notes to CRM records
Flow:
- Search transcripts for client meetings
- Extract summary, action items, and key quotes
- Create note on Attio company record with meeting details
技能: transcript-search → attio-crm
使用场景: 将会议笔记添加到CRM记录中
流程:
- 搜索客户会议的记录
- 提取摘要、行动项和关键引用
- 在Attio公司记录中创建包含会议详情的笔记