Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add ddpie/lark-mcp-on-agentcore --skill lark-basegit clone --depth 1 https://github.com/ddpie/lark-mcp-on-agentcoreWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/ddpie/lark-mcp-on-agentcore/lark-base)<a href="https://agentmods.dev/skills/ddpie/lark-mcp-on-agentcore/lark-base"><img src="https://agentmods.dev/badge/skills/ddpie/lark-mcp-on-agentcore/lark-base/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ddpie/lark-mcp-on-agentcore/lark-base"><img src="https://agentmods.dev/badge/skills/ddpie/lark-mcp-on-agentcore/lark-base.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00082 | $0.04998 |
| Opus 5 | $0.00041 | $0.02499 |
| Sonnet 5 | $0.00016 | $0.01000 |
| Haiku 4.5 | $0.00008 | $0.00500 |
Grade A, and why
lark-base scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 151 lines — stays where its author put it; the contents beside it link to each section on GitHub.
base
何时使用
使用本 skill：
- 用户明确提到 Base / 多维表格 / bitable，或给出
/base/链接。 - 用户要在 Base 内建表、改表、管理字段、写记录、查记录、配视图。
- 用户要在 Base 内做公式字段、lookup 字段、跨表计算、派生指标、筛选聚合、TopN、统计分析。
- 用户要管理 Base 表单、仪表盘、workflow、高级权限或角色。
- 用户要把旧 Base 聚合式命令或旧写法迁移到当前 shortcut。
不要使用本 skill：
- 只是认证、初始化配置、切换身份、处理 scope 或权限授权恢复（MCP server 自动处理认证）。
- 把本地 Excel / CSV /
.base导入成 Base，转lark_drive_import(type="bitable")。 - 泛化数据分析、字段设计、公式讨论，但没有 Base/多维表格上下文。
使用边界
- Base 业务操作只使用
lark_base_*工具，不使用旧聚合式命令。 - 本轮 Base 不依赖
lark_discover。SKILL 只保留路由、风险和复杂 JSON/DSL；简单命令由命令自身的参数、tips 和错误恢复承接。 - 用户要把 Excel / CSV /
.base导入成 Base 时，先转lark_drive_import(type="bitable")，导入完成后再回到 Base 工具。 - 认证由 MCP server 自动处理；Base 文档只保留会影响 Base 路径选择的权限规则。
先获取 Base Token 和所需 ID
进入任何需要目标 Base 的工具前，必须先拿到可用的 base_token，以及当前任务需要的 table_id / view_id / record_id / form_id / dashboard_id / workflow_id 等真实 ID；不要把完整 URL、wiki token、workspace token 或孤立 raw token 直接当作 base_token。
- 用户输入 URL 或分享链接：先调用
lark_base_url_resolve(url="<url>")，用返回的base_token和相关 ID 继续后续命令。 - 用户输入 Base 标题、关键词或不确定名称：先调用
lark_base_title_resolve(title="<keyword>")；title传入标题中的短关键词，不超过 30 个字符；过长标题先取最有区分度的短关键词；多候选时先让用户消歧，不要猜。 - 文档嵌入 Base 标签：直接读取
<bitable>/<base_refer>的token作为base_token，table-id作为table_id，view-id作为view_id；孤立 raw token 不走lark_base_url_resolve()。 - 仍无法定位且用户不是要新建 Base 时，先反问用户要操作哪一个 Base；用户要新建时才用
lark_base_base_create()。
快速路由
| 用户目标 | 优先工具 | 何时读 reference |
|---|---|---|
| 查 Base 本体 | lark_base_base_get() |
用返回确认 Base 名称、owner、权限和可继续操作的 token |
| 创建/复制 Base | lark_base_base_create() / lark_base_base_copy() |
新建时强烈推荐用 table_name + fields 同时配置新 Base 里唯一一个初始数据表的 name 和 schema；写入后报告新 Base 标识和 permission_grant |
| 查看 Base 内资源目录 | lark_base_base_block_list() |
想先了解一个 Base 里有哪些 table/docx/dashboard/workflow/folder 时优先用它；返回 ID 关系和 fewshot 看 lark_discover(query="base.base-block.list") |
| 管理 Base 内资源目录 | lark_base_base_block_create/move/rename/delete |
创建或整理 Base 直接管理的 folder/table/docx/dashboard/workflow；资源内容继续用对应工具 |
| 管理数据表 | lark_base_table_list/get/create/update/delete |
处理 table 的列出、详情、创建、重命名和删除 |
| 列/查/删字段 | lark_base_field_list/get/delete/search_options |
写入前用 list/get 确认字段类型、选项、ID；删除前确认目标字段 |
| 创建/更新字段 | lark_base_field_create() / lark_base_field_update() |
必读 lark_get_skill(domain="base", section="field-json")；公式读 lark_get_skill(domain="base", section="formula-field-guide")；lookup 读 lark_get_skill(domain="base", section="lookup-field-guide")；命令细节读 lark_get_skill(domain="base", section="field-create") / lark_get_skill(domain="base", section="field-update") |
| 读记录明细 | lark_base_record_get() / lark_base_record_list() / lark_base_record_search() |
涉及筛选、排序、Top/Bottom N、聚合、多表关联、全局结论时读 lark_get_skill(domain="base", section="data-analysis-sop") |
| 写记录 | lark_base_record_upsert() / lark_base_record_batch_create() / lark_base_record_batch_update() |
必读 lark_get_skill(domain="base", section="record-upsert") / lark_get_skill(domain="base", section="record-batch-create") / lark_get_skill(domain="base", section="record-batch-update") 和 lark_get_skill(domain="base", section="cell-value") |
| 附件字段 | lark_base_record_upload_attachment() / lark_base_record_download_attachment() / lark_base_record_remove_attachment() |
附件不要伪造成普通 CellValue；上传走本地文件，下载/删除按 file token 或字段定位 |
| 删除记录 / 分享记录链接 / 历史 | lark_base_record_delete() / lark_base_record_share_link_create() / lark_base_record_history_list() |
删除前确认 record；分享链接最多 100 条；历史读 lark_get_skill(domain="base", section="record-history-list")，只查单条记录，不做整表审计 |
| 管理视图 | lark_base_view_* |
lark_base_view_set_filter() 读 lark_get_skill(domain="base", section="view-set-filter")；其余配置先 get 现状，再按返回结构更新 |
| 一次性聚合统计 | lark_base_data_query() |
必读 lark_get_skill(domain="base", section="data-analysis-sop") 和入口 lark_get_skill(domain="base", section="data-query-guide")；完整 DSL 再读 lark_get_skill(domain="base", section="data-query") |
| 公式字段 | lark_base_field_create(json='{"type":"formula",...}') |
必读 lark_get_skill(domain="base", section="formula-field-guide")，读后再加隐藏确认 flag i_have_read_guide=true |
| Lookup 字段 | lark_base_field_create(json='{"type":"lookup",...}') |
必读 lark_get_skill(domain="base", section="lookup-field-guide")，读后再加隐藏确认 flag i_have_read_guide=true |
| 表单提交 | lark_base_form_submit() |
先读 lark_get_skill(domain="base", section="form-detail") 获取题目、filter 和附件所需 base_token；提交 JSON 读 lark_get_skill(domain="base", section="form-submit") |
| 表单题目创建/更新 | lark_base_form_questions_create() / lark_base_form_questions_update() |
读 lark_get_skill(domain="base", section="form-questions-create") / lark_get_skill(domain="base", section="form-questions-update") |
| 其他表单管理 | lark_base_form_list/get/detail/create/update/delete / lark_base_form_questions_list/delete |
lark_base_form_detail() 读 lark_get_skill(domain="base", section="form-detail")；删除前确认目标表单 |
| 仪表盘与组件 | lark_base_dashboard_* / lark_base_dashboard_block_* |
提到图表/看板/block 时先读 lark_get_skill(domain="base", section="dashboard")；组件 data_config 读 lark_get_skill(domain="base", section="dashboard-block-data-config")；读取图表计算结果用 lark_base_dashboard_block_get_data() |
| Workflow | lark_base_workflow_* |
创建/更新或理解 steps 时读入口 lark_get_skill(domain="base", section="workflow-guide") 和 steps JSON SSOT lark_get_skill(domain="base", section="workflow-schema")；list/get/enable/disable 只处理 workflow ID 与启停状态 |
| 高级权限与角色 | lark_base_advperm_* / lark_base_role_* |
角色操作先读入口 lark_get_skill(domain="base", section="role-guide")；角色 create/update 或解读完整配置再读权限 JSON SSOT lark_get_skill(domain="base", section="role-config")；系统角色不可删除；关闭高级权限会影响自定义角色 |
What ships with it
25 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/dashboard-block-data-config.md 12 KB
- references/formula-field-guide.md 54 KB
- references/lark-base-cell-value.md 5.2 KB
- references/lark-base-dashboard-block-get-data.md 23 KB
- references/lark-base-dashboard.md 10 KB
- references/lark-base-data-analysis-sop.md 11 KB
- references/lark-base-data-query-guide.md 2.8 KB
- references/lark-base-data-query.md 19 KB
- references/lark-base-field-create.md 3.3 KB
- references/lark-base-field-json.md 13 KB
- references/lark-base-field-update.md 6.8 KB
- references/lark-base-form-detail.md 2.9 KB
- references/lark-base-form-questions-create.md 4.0 KB
- references/lark-base-form-questions-update.md 2.4 KB
- references/lark-base-form-submit.md 5.9 KB
- references/lark-base-record-batch-create.md 1.9 KB
- references/lark-base-record-batch-update.md 1.9 KB
- references/lark-base-record-history-list.md 1.4 KB
- references/lark-base-record-upsert.md 2.1 KB
- references/lark-base-role-guide.md 3.0 KB
- references/lark-base-view-set-filter.md 3.8 KB
- references/lark-base-workflow-guide.md 27 KB
- references/lark-base-workflow-schema.md 38 KB
- references/lookup-field-guide.md 18 KB
- references/role-config.md 21 KB
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 151 lines · 82 tokens per session scan A 2a5b359f8188
lark-base is a skill published in the GitHub repository ddpie/lark-mcp-on-agentcore (8 stars, last pushed 11d ago), licensed MIT. It adds 82 tokens to every session and 4,998 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
bitable
Access Lark Bitable databases - list tables, view fields, and read records. Use when user asks about a Bitable, database, or wants to query structured data.
django-migration-psql
Reviews Django migration files for PostgreSQL best practices specific to Prowler. Trigger: When creating migrations, running makemigrations/pgmakemigrations, reviewing migration PRs, adding indexes or constraints to database tables, modifying existing migration files, or writing data backfill migrations. Always use…
postgresql-indexing
PostgreSQL indexing best practices for Prowler: index design, partial indexes, partitioned table indexing, EXPLAIN ANALYZE validation, concurrent operations, monitoring, and maintenance. Trigger: When creating or modifying PostgreSQL indexes, analyzing query performance with EXPLAIN, debugging slow queries, reviewing…
amazon aurora dsql
Deprecated compatibility redirect for Aurora DSQL guidance. Use when a request concerns DSQL, Aurora DSQL, distributed SQL, DSQL schemas, migrations, queries, authentication, performance, or application development.
dynamodb
AWS DynamoDB NoSQL database for scalable data storage. Use when designing table schemas, writing queries, configuring indexes, managing capacity, implementing single-table design, or troubleshooting performance issues.
rds
AWS RDS relational database service for managed databases. Use when provisioning databases, configuring backups, managing replicas, troubleshooting connectivity, or optimizing performance.