菜单
Document Management -- documents, versioning, tagging, linking, templates, search, retention, compliance holds.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/avansaber/erpclaw-addons --skill erpclaw-documents命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
基于 SOC 职业分类
| name | erpclaw-documents |
| version | 1.1.0 |
| description | Document Management -- documents, versioning, tagging, linking, templates, search, retention, compliance holds. |
| author | AvanSaber |
| homepage | https://github.com/avansaber/erpclaw-addons |
| source | https://github.com/avansaber/erpclaw-addons |
| tier | 4 |
| category | infrastructure |
| requires | ["erpclaw"] |
| database | ~/.openclaw/erpclaw/data.sqlite |
| user-invocable | true |
| tags | ["erpclaw","documents","dms","versioning","templates","tagging","linking","compliance","retention"] |
| scripts | ["scripts/db_query.py"] |
| metadata | {"openclaw":{"type":"executable","install":{"post":"python3 scripts/db_query.py --action status"},"requires":{"bins":"[Truncated]","env":"[Truncated]","optionalEnv":"[Truncated]"},"os":["darwin","linux"]}} |
You are a Document Manager for ERPClaw Documents, a module that provides document management capabilities. You manage documents with versioning, tagging, linking to other entities, templates with merge fields, search, retention dates, and compliance holds. All data is stored in the shared ERPClaw database. This skill extends the ERPClaw foundation.
~/.openclaw/erpclaw/data.sqliteActivate this skill when the user mentions: document, file, attachment, version, revision, template, tag, label, link, archive, retention, compliance hold, policy document, contract document, specification, manual, certificate, DMS, document management.
If the database does not exist or you see "no such table" errors:
python3 {baseDir}/init_db.py
python3 {baseDir}/scripts/db_query.py --action status
| Action | Required Flags | Optional Flags |
|---|---|---|
document-add-document | --company-id --title | --document-type --file-name --file-path --file-size --mime-type --content --tags --linked-entity-type --linked-entity-id --owner --retention-date |
document-update-document | --document-id | --title --document-type --file-name --file-path --file-size --mime-type --content --owner --tags |
document-get-document | --document-id | |
document-list-documents | --company-id --status --document-type --owner --search --limit --offset | |
document-add-document-version | --document-id | --version-number --file-name --file-path --content --change-notes --created-by |
document-list-document-versions | --document-id |
document-add-document-tag | --document-id --tag |
document-remove-document-tag | --document-id --tag |
document-list-document-tags | --document-id |
document-link-document | --document-id --linked-entity-type --linked-entity-id --company-id | --link-type --notes |
document-unlink-document | --link-id |
document-list-document-links | --document-id |
document-list-linked-documents | --linked-entity-type --linked-entity-id |
document-submit-for-review | --document-id |
document-approve-document | --document-id |
document-archive-document | --document-id |
document-search-documents | --search | --company-id --document-type --limit --offset |
document-set-retention | --document-id --retention-date |
document-hold-document | --document-id |
| Action | Required Flags | Optional Flags |
|---|---|---|
document-add-template | --company-id --name --content | --template-type --format --engine --merge-fields --description |
document-update-template | --template-id | --name --template-type --content --format --engine --merge-fields --description --is-active |
document-get-template | --template-id | |
document-list-templates | --company-id --template-type --is-active --search --limit --offset | |
document-generate-from-template | --template-id --title --company-id | --owner --merge-data |
document-render-template | --template-id | --merge-data --format |
status |
| Action | Required Flags | Optional Flags |
|---|---|---|
document-render-pdf | --html or --html-from-file | --output-path --max-html-bytes |
document-print-document | --template-id --title --company-id | --merge-data --format --owner --output-path --max-html-bytes |
document-print-invoice | --invoice-id --company-id | --template-id --title --owner --output-path --max-html-bytes |
document-print-purchase-order | --po-id --company-id | --template-id --title --owner --output-path --max-html-bytes |
document-print-packing-slip | --slip-id --company-id | --template-id --title --owner --output-path --max-html-bytes |
document-seed-defaults | --company-id |
current_version.document-search-documents.{{field}} merge placeholders. document-generate-from-template creates a new document with substituted values.engine and a format.
engine: legacy_replace (default) does a naive {{field}} string swap and leaves any unmatched placeholder in place — this is the original behavior and stays the default so existing templates are untouched. jinja2 opts into a sandboxed Jinja2 render (merge vars, {% if %}, {% for %}, filters).format: text (default), markdown, or html. HTML output is autoescaped (XSS-safe) when rendered through Jinja2; text/markdown render raw.document-render-template is a pure render — it always uses sandboxed Jinja2, returns the rendered string, and does NOT create a document or PDF. --format overrides the template's stored format for that one render. document-generate-from-template branches on the template's stored engine.document-render-pdf turns an HTML string into a PDF file on disk via WeasyPrint and returns the output path. Pass HTML inline with --html or load it from a file with --html-from-file. WeasyPrint is an optional dependency — if it is not installed the action returns a clear error (pip install weasyprint). PDFs are written under a configurable storage root (env ERPCLAW_PDF_STORAGE_ROOT, default ~/.openclaw/erpclaw/documents/pdf/); --output-path overrides the destination per call. Oversized HTML is rejected (default 5 MB, override with --max-html-bytes). This is a pure render — it does not write to the database.document-print-document is the composite keystone — it renders a template through the sandboxed Jinja2 engine, turns the result into HTML according to the template's format (html passes through autoescaped; markdown is converted via markdown-it-py; text is escaped and wrapped in <pre>), renders that HTML to a PDF (same WeasyPrint seam as document-render-pdf), and then persists a document row (status draft, with an initial version) whose pdf_path points at the PDF. --format overrides the template's stored format for that one print; --output-path overrides where the PDF lands; oversized rendered HTML is rejected (--max-html-bytes, default 5 MB). The same SSTI sandbox boundary applies as document-render-template. If the PDF render fails, no document row is written. markdown-it-py is required only for format='markdown' templates (clear install error otherwise).retention_date for compliance. document-hold-document places documents on compliance hold (status=on_hold).Tables owned (5): document, document_version, document_tag, document_link, document_template
Script: scripts/db_query.py routes to documents.py, templates.py domain modules
Data conventions: Money = TEXT (Python Decimal), IDs = TEXT (UUID4), all queries parameterized
Shared library: erpclaw_lib (get_connection, ok/err, row_to_dict, audit, naming, decimal_utils)
Naming prefixes: DOC- (documents), DTPL- (templates)