my_stock_log_skill

当且仅当用户明确提到“分析日志多维表”或“my_stock_log_skill”时触发。专门用于美股分析报告结论在钉钉多维表（dingtalk-ai-table）中的结构化归档（新增）与检索（查询）。

Security Scan

The skill's declared purpose (archiving and querying US-stock analysis rows in a DingTalk multidimensional table) matches its instructions and requested footprint; it is an instruction-only wrapper that delegates actual API access to the dingtalk-ai-table skill.

Purpose & Capability

Name, description, and runtime instructions all describe the same function (insert/query records in a specific dingtalk-ai-table). The skill does not request unrelated binaries, env vars, or file paths, and its hard-coded Base ID and Sheet ID are consistent with a single-table archiver.

Instruction Scope

SKILL.md restricts behavior to field validation, asking the user for missing data, delegating inserts/reads to the dingtalk-ai-table skill, and performing local filtering/presentation. This is within scope, but the skill relies entirely on another skill (dingtalk-ai-table) to perform network/API actions — you should verify that that downstream skill only performs the expected DingTalk operations and has appropriate permissions. The instructions are otherwise specific (field formats, truncation rules).

Install Mechanism

Instruction-only skill with no install spec and no code files; nothing is written to disk or downloaded during install.

Credentials

The skill itself requests no environment variables or credentials, which is proportionate. However, because it delegates to dingtalk-ai-table for API access, any credentials required to access DingTalk will be requested/used by that other skill — verify that those credentials are scoped to the intended table and not overly broad.

Persistence & Privilege

always is false and the skill is user-invocable; it does not request permanent/always-on presence or modify other skills' configs. No elevated persistence privileges are requested.

Assessment

This skill is a lightweight instruction wrapper that validates four fields and delegates reads/writes to the dingtalk-ai-table skill. Before installing, check the following: (1) confirm the downstream dingtalk-ai-table skill is trustworthy and you understand which DingTalk credentials it will use and their scope (access to the listed Base ID/Sheet ID); (2) verify the hard-coded Base ID and Sheet ID are the table you expect and that writing records there is acceptable; (3) test inserts/queries with non-sensitive sample data to confirm the delegation and field-validation behave as documented; and (4) consider data retention and access controls on the DingTalk table (sensitive analysis could be visible to others). If you are satisfied with the dingtalk-ai-table skill's permissions and the table target, the my_stock_log_skill itself is coherent and low-risk.

my_stock_log_skill (股票分析日志管理)

核心定位

此技能是分析报告归档与检索的调度中枢。它不直接操作钉钉 API，而是通过调用底层的 dingtalk-ai-table 技能来完成多维表的读写操作。

触发条件

• 只有当用户明确提到“分析日志多维表”或技能名称“my_stock_log_skill”时，才使用此技能。

目标数据源

• Base ID: 1OQX0akWmxpyBpnaCQQQoYkY8GlDd3mE
• Sheet ID: Shule1J

表格字段规范 (Schema)

在执行“新增记录”前，必须严格校验以下 4 个字段是否全部收集完毕且符合格式要求：

1. 标的: 美股标的简码（必须自动转为大写，如 AAPL）。
2. 分析时间: 时间格式必须为 YYYYMMDD_HHMMSS（年月是_时分秒，如 20260325_123000）。
3. 分析结论: 报告核心结论（使用英文大写，如 BUY、SELL、HOLD、STRONG BUY、STRONG SELL）。
4. 分析摘要: 总结内容的纯文本，必须限制在 300 字以内。如超出，须先由 AI 浓缩提炼。

核心能力与工作流

1. 新增记录 (Insert)

• 拦截校验：检查上述 4 个字段是否缺失或格式不符。如果有缺失，挂起操作并询问用户要求补充。绝不能插入“脏数据”。
• 执行新增：在尾部追加新记录。构建形如 {"标的": "AAPL", "分析时间": "20260325_123000", ...} 的 fields 结构，并交由 dingtalk-ai-table 提供的脚本执行 API 新增。

2. 精细查询 (Query)

• 多条件过滤：解析用户的自然语言（例如“查询特斯拉的BUY报告”），提取出查询维度（如标的=TSLA，结论=BUY）。
• 执行查询：通过 dingtalk-ai-table 获取该表的所有记录，并在本地按提取的条件进行精确或模糊匹配过滤。
• 展示结果：将过滤后的结果以清晰的 Markdown 表格形式展示给用户。