erpclaw-documents

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":["python3"],"env":[],"optionalEnv":["ERPCLAW_DB_PATH"]},"os":["darwin","linux"]}}

erpclaw-documents

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.

Security Model

• Local-only: All data stored in ~/.openclaw/erpclaw/data.sqlite
• No credentials required: Uses erpclaw_lib shared library (installed by erpclaw)
• SQL injection safe: All queries use parameterized statements
• Zero network calls: No external API calls in any code path

Skill Activation Triggers

Activate 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.

Setup (First Use Only)

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

Actions (Tier 1 -- Quick Reference)

Documents (19 actions)

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

Templates (7 actions)

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

Print / PDF (6 actions)

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

Key Concepts (Tier 2)

• Document Lifecycle: draft -> review -> approved -> published/archived. Can also be placed on_hold at any stage.
• Versioning: Each document tracks version history. Adding a version auto-increments current_version.
• Tagging: Free-form tags for categorization. Tags are also searched by document-search-documents.
• Linking: Documents can be linked to any entity (sales_order, customer, project, etc.) with link types: attachment, reference, supporting, supersedes.
• Templates: Reusable content with {{field}} merge placeholders. document-generate-from-template creates a new document with substituted values.
• Template engine / format: Each template carries an 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.
• Print / PDF: 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.
• Print a template to PDF + document: 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: Set retention_date for compliance. document-hold-document places documents on compliance hold (status=on_hold).
• Document Types: general, contract, policy, report, invoice, receipt, certificate, specification, manual, other.

Technical Details (Tier 3)

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)