← Back to Skills Marketplace
cgxxxxxxxxxxxx

Transcript Analysis

by cgxxxxxxxxxxxx · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
50
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install transcript-analysis
Description
从 Stock Analysis 或 QVeris API 抓取 earnings call transcript,进行主题信号挖掘,输出 evidence_ledger + theme_timeseries + summary_report。 触发词:分析业绩会、earnings call 分析、transcr...
README (SKILL.md)

Transcript Analysis Skill

从财报电话会逐字稿中挖掘主题信号,支持跨季度对比分析。


工作流程(Step 1-5)

Step 1: 确认分析参数

输入

  • 股票代码(如 LMT, PM)
  • 季度列表(如 2025Q4, 2025Q3)
  • 数据源模式(web 免费 / api QVeris)

输出:参数配置确认

✅ 检查点:股票代码与季度格式正确


Step 2: 🔍 参数确认检查点

输入

  • 用户提供的分析请求
  • 当前日期(用于默认输出目录命名)
Transcript 分析配置:
- 股票代码:LMT
- 季度:2025Q4, 2025Q3, 2025Q2
- 数据源:web(免费)
- 输出目录:./transcript_LMT_20260527/

确认继续?[Y/n/调整]

输出:用户确认的配置参数

✅ 检查点:用户明确确认后执行


Step 3: 执行分析脚本

输入

  • 确认后的配置参数
  • transcript_analyzer.py脚本路径
# 默认 web 模式(免费)
python transcript_analyzer.py --symbol LMT --quarters 2025Q4 --mode web

# 多季度对比
python transcript_analyzer.py --symbol LMT --quarters 2025Q4,2025Q3,2025Q2 --mode web

# API fallback
python transcript_analyzer.py --symbol PM --quarters 2025Q4 --mode api

输出

  • 脚本执行状态
  • 生成的CSV和Markdown文件

✅ 检查点:验证输出文件已生成


Step 4: 边界条件处理

输入

  • 脚本执行结果
  • 错误信息(如有)
场景 触发条件 处理动作
脚本不存在 transcript_analyzer.py 未找到 提示检查路径或重新创建
Stock Analysis 页面结构变化 解析失败 更新解析逻辑或切换 api 模式
QVeris 配额不足 API 返回 402 切换 web 模式或提示充值
无该季度 transcript 网页 404 跳过该季度,继续其他
网络超时 请求 > 30s 重试 3 次后切换数据源

输出:处理后的结果或错误提示

✅ 检查点:所有错误已妥善处理


Step 5: 输出文件

输入

  • 脚本生成的输出文件
文件 说明
{symbol}_evidence_ledger.csv 核心:每条信号追踪到原文片段、说话人、分段、语境
{symbol}_theme_timeseries.csv 跨季度主题热度(提及次数 + 每千词提及率)
{symbol}_summary_report.md 可读报告:主题排名 + 关键信号清单

输出:文件路径确认

✅ 检查点:用户确认输出文件位置


数据源优先级

优先级 数据源 成本 覆盖范围
1 Stock Analysis (stockanalysis.com) 免费 多数美股
2 QVeris API (Alpha Vantage) 2 credits/次 更广覆盖

参数说明

参数 必填 说明
--symbol 股票代码,如 LMT, PM
--quarters 季度列表逗号分隔,如 2025Q4,2025Q3
--mode 数据源: web(默认免费) / api(QVeris)
--output-dir 输出目录,默认当前目录

主题词库(24 个主题)

消费/烟草(PM 等): 无烟转型 / 尼古丁袋 / 电子烟 / 定价 / 销量份额 / 利润率 / 监管税收 / 创新 / 汇率 / 股东回报 / 成本重组 / 业绩指引 / 区域表现 / 竞争格局

军工/Defense(LMT 等): 积压订单 / 战斗机F-35 / 导弹弹药 / 太空业务 / 国防预算 / 多年期合同 / AI自主技术 / 金穹导弹防御 / 供应链产能 / 国际销售


输出目录惯例

./transcript_{symbol}_YYYYMMDD/

脚本位置

本技能捆绑 transcript_analyzer.py 脚本,安装后位于 skill 目录内。

Usage Guidance
Review before installing. Use only after the publisher removes and rotates the embedded QVeris key, changes the script to read a user-provided key from an environment variable or secret store, and clearly documents when authenticated API mode sends requests to QVeris. The default web mode is lower risk, but the shipped secret should still be treated as exposed.
Capability Tags
requires-sensitive-credentials
Capability Assessment
Purpose & Capability
Fetching earnings call transcripts and writing analysis reports is coherent with the stated purpose, but embedding a vendor API key in the distributed script is not necessary for that purpose and creates credential and billing exposure.
Instruction Scope
The skill documents web and API modes and asks for confirmation in the workflow, but the actual API credential is preloaded in code rather than requiring user-supplied configuration or explicit credential setup.
Install Mechanism
No separate installer, persistence hook, dependency install, or background setup behavior was found; the package contains SKILL.md and a Python script.
Credentials
Network access to Stock Analysis and QVeris and local report writes fit the tool, but authenticated API calls using a bundled secret are overbroad and under-controlled for a shared skill artifact.
Persistence & Privilege
The script creates an output directory and writes CSV/Markdown reports, with no observed privilege escalation or long-running persistence; the main privilege concern is the exposed API secret.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install transcript-analysis
  3. After installation, invoke the skill by name or use /transcript-analysis
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of transcript-analysis. - Supports earnings call transcript theme signal mining via Stock Analysis or QVeris API. - Step-by-step workflow from parameter confirmation through results output, with clear checkpoints. - Robust handling for various edge cases, including data source switching and error prompts. - Outputs: evidence ledger, theme timeseries, and summary report files for multi-quarter analysis. - Includes topic dictionary for consumer/tobacco and defense sectors.
Metadata
Slug transcript-analysis
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Transcript Analysis?

从 Stock Analysis 或 QVeris API 抓取 earnings call transcript,进行主题信号挖掘,输出 evidence_ledger + theme_timeseries + summary_report。 触发词:分析业绩会、earnings call 分析、transcr... It is an AI Agent Skill for Claude Code / OpenClaw, with 50 downloads so far.

How do I install Transcript Analysis?

Run "/install transcript-analysis" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Transcript Analysis free?

Yes, Transcript Analysis is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Transcript Analysis support?

Transcript Analysis is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Transcript Analysis?

It is built and maintained by cgxxxxxxxxxxxx (@cgxxxxxxxxxxxx); the current version is v1.0.0.

💬 Comments