← 返回 Skills 市场
richard052317

Bi2 Report Generator

作者 Richard052317 · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
63
总下载
0
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install bi2-report-generator
功能描述
根据PPTX截图提取经营数据，自动生成并保存BI2事业部多维度HTML格式经营分析报告，并发送至飞书用户。
使用说明 (SKILL.md)

BI2 事业部经营分析报告生成器

用途

根据 PPTX 截图数据自动生成 BI2 事业部经营分析报告（HTML 格式）。

触发指令

  • "生成 BI2 经营分析报告"
  • "BI2 报告"
  • "事业部经营分析"
  • "BI2 经营分析"

输入

  • PPTX 截图（包含销售额、毛利额、毛利率、费用、客户、产品、库存、应收账款、未来预测等数据）
  • 用户标识（飞书用户，用于确定保存路径）

输出

  • HTML 格式经营分析报告
  • 保存路径：/Users/zhangzhaoliangdemacmini/Desktop/Openclaw_Pool/{USER}_Folder/BI2 事业部经营分析报告_{YYYYMMDD}.html

报告维度（10 个）

  1. 经营摘要：核心经营亮点、KPI 卡片、2026 年累计数据
  2. 收入分析：季度对比、月度销售额/毛利额/毛利率对比
  3. 盈利分析：毛利额趋势、净利润/净利率趋势、毛利差异因素分析
  4. 费用分析：季度费用对比、月度费用明细、费用洞察
  5. 客户分析：2025 年客户排名、2026 年累计客户排名
  6. 产品分析：HT 固态产品线、EVP 液态产品线、产品策略建议
  7. 库存分析：库存 KPI 卡片、库存改善说明
  8. 应收账款：应收账款概况、重要说明（逾期规则 + 呆滞库存）、客户欠款排序
  9. 未来预测：4-6 月预测数据、预算差距警示、预测洞察
  10. 洞察建议：核心经营建议、销售策略、产品路线图、费用管控

处理流程

  1. 读取截图：使用 image 工具读取所有截图，提取紫色底色实际数据
  2. 数据校验：检查数据完整性（销售额、毛利额、毛利率、净利润等）
  3. 生成报告：使用 HTML 模板填充数据
  4. 保存文件：根据飞书用户确定保存路径
  5. 发送飞书：上传 HTML 文件并发送消息
  6. 清理临时文件：删除临时脚本

数据规范

金额单位

  • 统一使用 万元
  • 2025 年全年销售额：115,496 万元（11.55 亿元）

数据格式

  • 销售额：整数（如 11,191）
  • 毛利额：整数（如 2,108）
  • 毛利率：百分比（如 18.84%）
  • 净利润：整数（如 873）
  • 净利率：百分比（如 7.80%）

关键数据（基于截图紫色底色）

月份 销售额 毛利额 毛利率 净利润 净利率
2025-01 8,199 1,617 19.73% 48 0.58%
2025-02 7,922 1,546 19.52% 121 1.52%
2025-03 9,854 2,035 20.66% 1,378 13.98%
2025-04 11,046 2,572 23.29% 1,291 11.69%
2025-05 10,009 2,294 22.92% 1,882 18.81%
2025-06 10,474 2,427 23.17% 968 9.24%
2025-07 13,166 3,333 25.32% 1,662 12.62%
2025-08 11,392 2,734 24.00% 971 8.52%
2025-09 6,576 1,716 24.87% 453 6.88%
2025-10 5,766 1,434 24.60% 36 0.63%
2025-11 7,467 1,837 21.92% -60 -0.81%
2025-12 6,234 1,139 18.27% 1,094 17.55%
2026-01 8,716 1,537 17.63% 432 4.95%
2026-02 5,407 813 15.04% -251 -4.65%
2026-03 11,191 2,108 18.84% 873 7.80%

飞书用户映射

飞书用户 英文名 保存路径
Richard (张召良) Richard /Users/zhangzhaoliangdemacmini/Desktop/Openclaw_Pool/Richard_Folder/
JING (李京) JING /Users/zhangzhaoliangdemacmini/Desktop/Openclaw_Pool/JING_Folder/
Jeanie (王琴) Jeanie /Users/zhangzhaoliangdemacmini/Desktop/Openclaw_Pool/Jeanie_Folder/
Mickey (熊伟) Mickey /Users/zhangzhaoliangdemacmini/Desktop/Openclaw_Pool/Mickey_Folder/
City (袁思婷) City /Users/zhangzhaoliangdemacmini/Desktop/Openclaw_Pool/City_Folder/

文件名规范

  • 标准格式：BI2 事业部经营分析报告_{YYYYMMDD}.html
  • 多版本格式：BI2 事业部经营分析报告_{YYYYMMDD}_v{版本号}.html
  • 最终版本：BI2 事业部经营分析报告_{YYYYMMDD}_完善版.html

注意事项

  1. 数据准确性：必须基于截图紫色底色实际数据，不可编造
  2. 单位统一：所有金额统一为万元
  3. 中文显示：HTML 必须使用微软雅黑字体，避免乱码
  4. 交互友好：使用导航标签切换章节，图表可视化
  5. 版本管理：多轮修改时保留版本号，用户确认后清理中间版本

技能位置

~/.openclaw/workspace/skills/bi2-report-generator/

版本历史

  • v1.0 (2026-04-13)：初始版本，支持 10 个维度报告生成
安全使用建议
Do not install or run this skill without changes and review. Specific issues to address before trusting it: 1) Remove hard-coded Feishu credentials from source and require the operator to supply app_id/app_secret (or a tenant token) via secure environment variables or a secrets manager; verify the credentials belong to an approved application. 2) Implement or verify the image-parsing step (SKILL.md promises screenshot extraction but generator.py uses static DEFAULT_DATA). 3) Make file save paths configurable instead of using hard-coded developer desktop paths. 4) Clarify how Feishu user open_ids are obtained and ensure the skill only sends files to intended recipients. 5) Audit network behavior (requests.post) and restrict the app's permissions to the minimum required. If you cannot verify the origin/ownership of the embedded Feishu app_secret or cannot run the code in a safe, isolated environment, avoid using this skill.
功能分析
Type: OpenClaw Skill Name: bi2-report-generator Version: 1.0.0 The skill bundle contains hardcoded sensitive credentials (Feishu/Lark `app_id` and `app_secret`) in `generator.py`, which allows unauthorized access to the associated Feishu application. Additionally, both `SKILL.md` and `generator.py` use hardcoded absolute file paths targeting a specific user's local directory (`/Users/zhangzhaoliangdemacmini/Desktop/Openclaw_Pool/`), which is highly irregular and poses a risk of unauthorized filesystem interaction. While the logic appears to align with the stated purpose of report generation, the inclusion of secrets and environment-specific paths suggests either a poorly sanitized internal tool or a potential vector for credential abuse.
能力评估
⚠ Purpose & Capability
The skill's stated purpose (extract data from PPTX screenshots and generate/send reports) is reasonable, but the code does not implement image extraction — it uses DEFAULT_DATA instead. The code also embeds a Feishu app_id/app_secret and writes to hard-coded user desktop paths (specific to one developer), which is disproportionate and unexpected for a generic report generator. Hard-coded tenant credentials and machine-specific paths are inconsistent with a portable, user-facing skill.
⚠ Instruction Scope
SKILL.md instructs the agent to use an image tool to read screenshots, validate extracted data, clean temporary scripts, and map Feishu users to save paths. The provided generator.py does not parse images, does not perform data extraction from screenshots, and does not delete temp files. It defines a send_to_feishu function that requires a user_open_id, but the SKILL.md's Feishu user mapping supplies local folder names, not open_ids — the runtime wiring is incomplete and unclear.
✓ Install Mechanism
No install spec is provided and the skill is instruction-plus-source only, so nothing external is downloaded or automatically installed. This lowers install-time risk.
⚠ Credentials
No environment variables are declared, yet generator.py contains hard-coded Feishu app_id and app_secret (sensitive credentials) used to obtain tenant tokens. Embedding tenant-level credentials in code is inappropriate and increases the risk of unauthorized access or exfiltration; credential handling is not proportional or transparent.
ℹ Persistence & Privilege
The skill writes files to and creates folders under hard-coded local paths (e.g., /Users/zhangzhaoliangdemacmini/Desktop/Openclaw_Pool/...). It does not request always:true and does not modify other skills. Writing to user folders is expected for a report generator, but the fixed, user-specific paths reduce portability and may leak developer-specific environment assumptions.
如何使用
  1. 确保已安装 OpenClaw（本地或 Docker 部署）
  2. 在对话框中输入安装命令：/install bi2-report-generator
  3. 安装完成后，直接呼叫该 Skill 的名称或使用 /bi2-report-generator 触发
  4. 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
BI2 事业部经营分析报告生成器 1.0.0 - 首次发布：自动根据 PPTX 截图数据生成 BI2 事业部经营分析报告（HTML 格式） - 支持 10 个经营分析维度，涵盖收入、盈利、费用、客户、产品、库存、应收账款、未来预测等 - 输出文件支持用户专属路径管理和多版本命名规范 - 全流程自动化：截图读取、数据校验、报告生成、飞书分享、清理临时文件 - 统一数据单位和格式，保障报告规范与展示质量
元数据
Slug bi2-report-generator
版本 1.0.0
许可证 MIT-0
累计安装 1
当前安装数 1
历史版本数 1
常见问题

Bi2 Report Generator 是什么？

根据PPTX截图提取经营数据，自动生成并保存BI2事业部多维度HTML格式经营分析报告，并发送至飞书用户。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 63 次。

如何安装 Bi2 Report Generator？

在 OpenClaw 或 Claude Code 对话框中运行命令「/install bi2-report-generator」即可一键安装，无需额外配置。

Bi2 Report Generator 是免费的吗？

是的，Bi2 Report Generator 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。

Bi2 Report Generator 支持哪些平台？

Bi2 Report Generator 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。

谁开发了 Bi2 Report Generator？

由 Richard052317（@richard052317）开发并维护，当前版本 v1.0.0。

推荐 Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191115 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 留言讨论