edalcin/biodiversitydataskills

grist-master

Provides comprehensive technical knowledge about Grist (getgrist.com), the spreadsheet-database hybrid tool: REST API, SQL endpoint, MCP server, OAuth apps, webhooks, Python formulas and the full Excel-like function reference, column types, references/looku…

查看源码

使用方法

  1. 复制安装命令,在自己的智能体环境安装前先检查源码包。
  2. 从原始文档核对所需工具、凭据和支持的宿主。
  3. 说明输入材料与期望交付物,并对照来源检查输出。
npx skills add https://github.com/edalcin/biodiversitydataskills --skill grist-master

开始前需了解

  • 安装不包含模型订阅、API 额度或访问文件的权限。
  • 仓库星数和安装量反映活跃情况,不等于独立质量或安全测试。
仓库原始内容

按源仓库内容呈现,保留标题、案例、代码、表格、链接以及原文引用的演示图片。

Grist Master

Reference knowledge base for Grist, the open-source spreadsheet-database hybrid. Built from the official Grist Help Center. Use this skill to answer questions, write formulas, call the API, configure self-hosted instances, or build integrations without re-deriving Grist behavior from scratch.

This is a pure reference skill — no scripts, no dependencies. Read the relevant file(s) below before answering non-trivial Grist questions.

What Grist is

Grist stores data as tables of records (not spreadsheet cells with row/column addresses). Each table has typed columns; formulas are Python, apply to a whole column at once (no per-row A1-style addressing), and are recalculated automatically. A document = one SQLite file (.grist), containing all tables, pages, widgets, and access rules.

Quick reference — where to look

TopicFile
REST API (orgs/workspaces/docs/records/tables/columns/attachments/webhooks), SQL endpoint, API keys, OAuth apps, client librariesreferences/API_REFERENCE.md
MCP server setup (hosted + self-hosted), OAuth/CIMD, tool cataloguereferences/MCP_SERVER.md
Formula basics, Python sandbox, trigger formulas, Record/RecordSet/UserTable, lookupOne/lookupRecords, full function reference (Date/Logical/Lookup/Math/Stats/Text), summary tablesreferences/FORMULAS_AND_FUNCTIONS.md
Column types (Text, Numeric, Integer, Toggle, Date, DateTime, Choice, Choice List, Reference, Reference List, Attachment), two-way references, dropdown filteringreferences/COLUMN_TYPES.md
Access rules: enabling, default rules, row/column restrictions, user/rec/newRec, permissions (R/U/C/D/S), link keys, user attribute tablesreferences/ACCESS_RULES.md
Self-hosted Grist: Docker install, sandboxing (gvisor), teams, auth, full-edition activation, storage/backends, snapshots, telemetry, upgradesreferences/SELF_HOSTED.md
Webhooks, Zapier/n8n/Make/Pabbly/Integrately, embedding (iframe), custom widget Plugin APIreferences/INTEGRATIONS.md
Frequently asked questions (accounts, plans, sharing, embedding, row limits)references/FAQ.md
Glossary of Grist UI/data-model termsreferences/GLOSSARY.md

Core concepts an agent must get right

  • Formulas are Python 3.11, evaluated in a network-isolated sandbox with no

persistent filesystem. $Field in a formula == rec.Field. A formula applies to the whole column, not one cell.

  • References, not VLOOKUP. Cross-table relationships use Reference /

Reference List columns (foreign keys to a whole record) or Table.lookupOne(...) / Table.lookupRecords(...). VLOOKUP exists only as a lookupOne alias for spreadsheet-familiarity.

  • Trigger formulas run once on new records and/or on every update of

specific columns; they see extra value/user variables and are how you implement timestamps, authorship, or data sanitization — not by using a regular formula.

  • Summary tables are Grist's GROUP BY — never hand-roll aggregate rows

in a data table; add a summary widget and use $group.

  • Access rules are evaluated top-to-bottom per rule group in order

(column rules → table rules → default rules); first matching rule wins for each permission letter. The S (structure) permission can bypass all other restrictions because formulas aren't sandboxed from data.

  • API auth: a personal API key carries full account access (Authorization: Bearer <key>

against docs.getgrist.com or <team>.getgrist.com). For third-party tools acting on a user's behalf, prefer OAuth apps (scoped, revocable) over handing out API keys.

  • SQL endpoint (/api/docs/{docId}/sql) is read-only SELECT

against the document's underlying SQLite — separate from the records/tables CRUD endpoints, and may not fully respect row-level access rules, so treat it as an owner-only tool.

  • Self-hosted: gristlabs/grist Docker image toggles between Community

(free, OSS, Apache-2.0) and the full edition (30-day trial, then an activation key). Always set GRIST_SANDBOX_FLAVOR=gvisor in production — it's what isolates Python formulas from the host and network.

When answering Grist questions

  1. Formula / function question → read FORMULAS_AND_FUNCTIONS.md first;

don't guess Excel-function availability, check the table there (many Excel-lookalikes like HLOOKUP, INDEX, MATCH are not implemented in Grist — lookupRecords/references are the idiomatic replacement).

  1. "Call the Grist API" → read API_REFERENCE.md for the exact path/verb and

auth header; distinguish REST CRUD endpoints from the SQL endpoint.

  1. "Connect an AI assistant to Grist" → MCP_SERVER.md, not the REST API.
  2. "Self-host" / Docker / env vars → SELF_HOSTED.md.
  3. "Who can see/edit this" → ACCESS_RULES.md.

按任务相关推荐

来自同一仓库

更多 Skills

全部 Skills
edalcin
社区

biohousekeeper

Analyzes biodiversity spreadsheets (CSV/XLSX) and proposes a restructured column layout aligned with Darwin Core (DwC), asking the user clarifying questions about anything ambiguous before finalizing the report. Detects columns that should be renamed to DwC terms, composite columns that should be split (e.g. packed "lat,long" pairs, "Genus species" binomials, delimited locality hierarchies), redundant duplicate columns, and missing recommended fields (occurrenceID, basisOfRecord, eventDate, coor

安装量
1
GitHub Stars
4
最近更新
8月26日
edalcin
社区

darwin-core

Helps users work with Darwin Core (DwC), Darwin Core Archive (DwC-A), Darwin Core Conceptual Model (DwC-CM), and Darwin Core Data Package (DwC-DP), the biodiversity data standards maintained by TDWG. Validates DwC-A files, generates templates, maps CSV columns to DwC terms, explains the standard, explains class relationships via DwC-CM, and helps create DwC-DP packages. Use when the user mentions "Darwin Core", "DwC", "DwC-A", "DwC-DP", "DwC-CM", "GBIF", "biodiversity data", "occurrence", "taxon

安装量
1
GitHub Stars
4
最近更新
8月26日
edalcin
社区

iczn

Expert knowledge of the International Code of Zoological Nomenclature (ICZN, 4th edition 1999, with Declarations 44-47 and the 2012 electronic-publication amendment) for zoological taxonomists and for anyone structuring taxonomic data. Decides whether a name is available, which of two competing names is valid, whether a type designation is effective, how authorship and parentheses are cited, and how to model names, taxa and nomenclatural acts in a database. Use this skill whenever the user menti

安装量
1
GitHub Stars
4
最近更新
8月26日
edalcin
社区

skos-xl

Helps users build, validate, convert, and explore controlled vocabularies using SKOS (Simple Knowledge Organization System) and SKOS-XL (the W3C extension for annotatable labels). Supports Darwin Core integration for biodiversity vocabularies (basisOfRecord, habitat types, taxonomic names) and Traditional Knowledge (CTA/EtnoTermos) vocabularies with CARE principles, Nagoya Protocol compliance, per-label access control, and indigenous language attribution (PROV-O). Use when the user mentions "SKO

安装量
1
GitHub Stars
4
最近更新
8月26日