Manage Apple Reminders via the `remindctl` CLI on macOS (list, add, edit, complete, delete)....
npx skills add anomalyarmor/agents --skill "armor-tags"
Install specific skill from multi-skill repository
# Description
Classify and govern data with tags. Handles "tag this table as PII", "apply financial tag", "list tags", "classify data", "add governance label".
# SKILL.md
name: armor-tags
description: Classify and govern data with tags. Handles "tag this table as PII", "apply financial tag", "list tags", "classify data", "add governance label".
hooks:
PreToolUse:
- matcher: "Bash"
hooks:
- type: command
command: "python ${CLAUDE_PLUGIN_ROOT}/scripts/ensure-auth.py"
once: true
Data Classification with Tags
Organize and classify database objects with tags for governance, compliance, and business categorization.
Prerequisites
- AnomalyArmor API key configured (
~/.armor/config.yamlorARMOR_API_KEYenv var) - Python SDK installed (
pip install anomalyarmor)
When to Use
- "Tag this table as PII"
- "Apply financial reporting tag"
- "List all tags for this asset"
- "Mark these columns as sensitive"
- "Classify this data as confidential"
- "Add governance labels"
Concepts
Tag Categories
- business: Business domain tags (e.g., "finance", "marketing", "sales")
- technical: Technical classification (e.g., "fact_table", "dimension", "staging")
- governance: Compliance and security (e.g., "pii", "confidential", "gdpr")
Tag Scope
Tags can be applied to:
- Tables: Full table classification
- Columns: Column-level classification (for PII, sensitive data)
Steps
Creating a Tag
- Identify the asset and object (table or column) to tag
- Choose the tag name and category
- Call
client.tags.create()with the object path
Applying Multiple Tags
- Prepare list of tag names to apply
- Prepare list of object paths
- Call
client.tags.apply()for batch operations
Bulk Tagging Across Assets
- Create tag name
- List asset IDs to tag
- Call
client.tags.bulk_apply()
Example Usage
List Existing Tags
from anomalyarmor import Client
client = Client()
# List all tags for an asset
tags = client.tags.list(asset="postgresql.analytics")
for tag in tags:
print(f" {tag.name} ({tag.category}): {tag.object_path}")
# Filter by category
governance_tags = client.tags.list(
asset="postgresql.analytics",
category="governance"
)
Tag a Table as PII
tag = client.tags.create(
asset="postgresql.analytics",
name="pii_data",
object_path="public.customers",
object_type="table",
category="governance",
description="Contains personally identifiable information"
)
print(f"Created tag: {tag.id}")
Tag a Column as Sensitive
tag = client.tags.create(
asset="postgresql.analytics",
name="sensitive",
object_path="public.customers.email",
object_type="column",
category="governance"
)
Apply Multiple Tags to Multiple Tables
result = client.tags.apply(
asset="postgresql.analytics",
tag_names=["financial_reporting", "quarterly_data"],
object_paths=["gold.fact_orders", "gold.fact_revenue", "gold.dim_customers"],
category="business"
)
print(f"Applied: {result.applied}, Failed: {result.failed}")
Tag Multiple Assets
result = client.tags.bulk_apply(
tag_name="production_critical",
asset_ids=["postgresql.analytics", "postgresql.warehouse", "snowflake.main"],
category="technical"
)
print(f"Tagged {result.applied} assets")
Expected Output
Tags for postgresql.analytics:
pii_data (governance): public.customers
financial_reporting (business): gold.fact_orders
quarterly_data (business): gold.fact_revenue
production_critical (technical): asset-level
By Category:
governance: 3 tags
business: 5 tags
technical: 2 tags
Follow-up Actions
- After tagging PII: Set up access controls and audit logging
- After business classification: Use tags to filter dashboards and reports
- After technical tagging: Use tags to prioritize monitoring
- To view tagged data: Filter assets by tag in the AnomalyArmor dashboard
# Supported AI Coding Agents
This skill is compatible with the SKILL.md standard and works with all major AI coding agents:
Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.