Agent-Skills.md

Agent Skills: Notion Master

Shared resource library for Notion integration skills. DO NOT load directly - this provides common references (setup, API docs, error handling, database schema) and scripts used by notion-connect, query-notion-db, import-skill-to-nexus, and export-skill-to-notion.

UncategorizedID: abdullahbeam/nexus-design-abdullah/notion-master
31

Install this agent skill to your local

pnpm dlx add-skill https://github.com/abdullahbeam/nexus-design-abdullah/tree/HEAD/00-system/skills/notion/notion-master

Skill Files

Browse the full folder contents for notion-master.

Download Skill

Loading file tree…

00-system/skills/notion/notion-master/SKILL.md

Skill Metadata

Name
notion-master
Description
Shared resource library for Notion integration skills. DO NOT load directly - this provides common references (setup, API docs, error handling, database schema) and scripts used by notion-connect, query-notion-db, import-skill-to-nexus, and export-skill-to-notion.

Notion Master

This is NOT a user-facing skill. It's a shared resource library referenced by Notion integration skills.

Purpose

Provides shared resources to eliminate duplication across:

  • notion-connect - Meta-skill for complete Notion workspace integration (NEW)
  • query-notion-db - Browse and search Notion skills database
  • import-skill-to-nexus - Download skills from Notion to local Nexus
  • export-skill-to-notion - Push skills to Notion for team sharing

Instead of loading this skill, users directly invoke the specific skill they need above.

Architecture: DRY Principle

Problem solved: The 3 Notion skills had 950 lines of duplicated content (setup wizard 3x, API docs 3x, error tables 3x).

Solution: Extract shared content into notion-master/references/ and notion-master/scripts/, then reference from each skill.

Result: 60% context reduction (950 → 370 lines in SKILL.md files)

Shared Resources

All 3 Notion skills reference these resources (progressive disclosure).

references/

setup-guide.md - Complete setup wizard

  • Getting Notion API key
  • Finding database ID
  • Configuring .env and user-config.yaml
  • Getting user's Notion ID

api-reference.md - Notion API patterns

  • Database query API
  • File upload API (3-step process)
  • File download API (3-step process)
  • Common curl examples

error-handling.md - Troubleshooting

  • Common errors and solutions
  • API error codes
  • Configuration issues
  • Network and timeout handling

database-schema.md - Beam Nexus Skills database

  • Property types and descriptions
  • Available Teams and Integrations
  • Field mapping table
  • Example entries

filter-syntax.md - Query filter syntax

  • Operators by property type
  • Common filter patterns
  • Sorting options

property-types.md - Property type reference

  • All 20+ property types
  • Read and write formats
  • Validation rules

block-types.md - Block type reference

  • All 50+ block types
  • Block schemas and examples
  • Rich text formatting

scripts/

Configuration & Setup

check_notion_config.py - Pre-flight validation

python check_notion_config.py [--json]

| Argument | Required | Default | Description | |----------|----------|---------|-------------| | --json | No | False | Output structured JSON for AI consumption |

Exit codes: 0=configured, 1=partial, 2=not configured

When to Use: Run this FIRST before any Notion operation. Use to validate API key is configured, check if Skills DB ID is set, diagnose connection issues, or determine if setup wizard is needed.

setup_notion.py - Interactive setup wizard

python setup_notion.py

No arguments - runs interactively. Guides through API key setup, tests connection, saves to .env, gets user's Notion ID, auto-runs database discovery.

When to Use: Use when Notion integration needs initial setup, when check_notion_config.py returns exit code 2, or when user needs to reconfigure credentials or change integration settings.

discover_databases.py - Database discovery (POST /search)

python discover_databases.py [--refresh] [--json] [--detailed]

| Argument | Required | Default | Description | |----------|----------|---------|-------------| | --refresh | No | False | Force refresh even if context file exists | | --json | No | False | Output JSON only (no file save) | | --detailed | No | False | Fetch full schema for each database (slower) |

Saves to: 01-memory/integrations/notion-databases.yaml

When to Use: Use when user asks "what databases do I have", "list my notion databases", after connecting a new integration, or when database name resolution fails (--refresh to update cache).

Database Querying

search_skill_database.py - Unified database querying (POST /databases/{id}/query)

# General mode
python search_skill_database.py --db "Database Name" [--filter EXPR] [--sort FIELD] [--sort-dir asc|desc] [--limit N] [--json]

# Skills DB preset mode
python search_skill_database.py --skills [--team TEAM] [--integration INT] [--name TEXT] [--owner ID] [--json]

| Argument | Required | Default | Description | |----------|----------|---------|-------------| | --db | No* | - | Database name or ID (fuzzy match) | | --skills | No* | False | Query Beam Nexus Skills database (preset mode) | | --team | No | - | [Skills mode] Filter by team (General, Solutions, etc.) | | --integration | No | - | [Skills mode] Filter by integration (Beam AI, Linear, etc.) | | --name | No | - | [Skills mode] Filter by skill name (partial match) | | --owner | No | - | [Skills mode] Filter by owner user ID | | --filter | No | - | Filter expression (see syntax below) | | --sort | No | - | Property name to sort by | | --sort-dir | No | desc | Sort direction: asc or desc | | --limit | No | - | Limit number of results | | --json | No | False | Output raw JSON |

*Either --db or --skills is required

Filter Syntax:

--filter "Status = Active"           # Equals
--filter "Status != Done"            # Not equals
--filter "Name contains project"     # Contains
--filter "Priority > 5"              # Greater than
--filter "Due Date < 2025-01-01"     # Less than
--filter "Tags is_empty"             # Is empty
--filter "Tags is_not_empty"         # Is not empty

When to Use: Use when user wants to query any Notion database, search for records, filter by properties, find skills in the Skills DB (--skills mode), or list database contents with sorting/pagination.

Page Operations

create_page.py - Create page in database (POST /pages)

python create_page.py --db "Database" [--properties JSON] [--interactive] [--json]

| Argument | Required | Default | Description | |----------|----------|---------|-------------| | --db | Yes | - | Database name or ID (fuzzy match) | | --properties | No | - | JSON object with property values | | --interactive | No | False | Prompt for each property | | --json | No | False | Output raw JSON |

When to Use: Use when user wants to add a new entry to a database, create a new record, or add an item to a Notion table. Use --interactive for guided property entry.

manage_page.py - Page management (GET/PATCH /pages/{id})

python manage_page.py get --page PAGE_ID [--json]
python manage_page.py update --page PAGE_ID --properties JSON [--json]
python manage_page.py delete --page PAGE_ID [--confirm] [--json]

Subcommand: get | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --page | Yes | - | Page ID | | --json | No | False | Output raw JSON |

Subcommand: update | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --page | Yes | - | Page ID | | --properties | Yes | - | JSON object with property values | | --json | No | False | Output raw JSON |

Subcommand: delete | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --page | Yes | - | Page ID | | --confirm | No | False | Skip confirmation prompt | | --json | No | False | Output raw JSON |

When to Use: Use get to retrieve page details/properties, update to modify page properties (status, tags, dates), delete to archive/remove pages. Requires page ID from query results.

Database Management

manage_database.py - Database operations (POST/PATCH /databases)

python manage_database.py create --parent PAGE_ID --title "Title" [--properties JSON] [--inline] [--json]
python manage_database.py update --db DB_ID [--title TEXT] [--add-property JSON] [--update-property JSON] [--json]

Subcommand: create | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --parent | Yes | - | Parent page ID | | --title | Yes | - | Database title | | --properties | No | [] | JSON array of property definitions | | --inline | No | False | Create inline database | | --json | No | False | Output raw JSON |

Subcommand: update | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --db | Yes | - | Database ID | | --title | No | - | New title | | --add-property | No | - | JSON property definition to add | | --update-property | No | - | JSON property update | | --json | No | False | Output raw JSON |

When to Use: Use create to make a new database inside a page, update to add/modify database properties (columns), change database title, or add new select options.

Block Operations

manage_blocks.py - Block operations (GET/PATCH/DELETE /blocks/{id})

python manage_blocks.py get --block BLOCK_ID [--json]
python manage_blocks.py children --block BLOCK_ID [--limit N] [--all] [--json]
python manage_blocks.py append --block BLOCK_ID [--content JSON] [--type TYPE] [--text TEXT] [--language LANG] [--checked] [--json]
python manage_blocks.py update --block BLOCK_ID --content JSON [--json]
python manage_blocks.py delete --block BLOCK_ID [--confirm] [--json]

Subcommand: get | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --block | Yes | - | Block ID | | --json | No | False | Output raw JSON |

Subcommand: children | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --block / --page | Yes | - | Block or page ID | | --limit | No | 50 | Max blocks to return | | --all | No | False | Fetch all pages | | --json | No | False | Output raw JSON |

Subcommand: append | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --block / --page | Yes | - | Parent block or page ID | | --content | No | - | JSON block content | | --type | No | - | Block type (paragraph, heading_1, etc.) | | --text | No | - | Text content | | --language | No | - | Code block language | | --checked | No | False | For to_do blocks | | --json | No | False | Output raw JSON |

Subcommand: update | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --block | Yes | - | Block ID | | --content | Yes | - | JSON block content | | --json | No | False | Output raw JSON |

Subcommand: delete | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --block | Yes | - | Block ID | | --confirm | No | False | Skip confirmation | | --json | No | False | Output raw JSON |

When to Use: Use get to retrieve block info, children to read page content, append to add text/headings/code blocks to pages, update to modify existing blocks, delete to remove content.

User Management

manage_users.py - User operations (GET /users)

python manage_users.py list [--limit N] [--all] [--save] [--json]
python manage_users.py get --user USER_ID [--json]
python manage_users.py me [--json]

Subcommand: list | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --limit | No | 100 | Max users to return | | --all | No | False | Fetch all pages | | --save | No | False | Save users to context file | | --json | No | False | Output raw JSON |

Subcommand: get | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --user | Yes | - | User ID | | --json | No | False | Output raw JSON |

Subcommand: me | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --json | No | False | Output raw JSON |

When to Use: Use list to see all workspace members, get to lookup a specific user, me to get current authenticated user/bot identity (useful for setting notion_user_id during setup).

Comment Operations

manage_comments.py - Comment operations (GET/POST /comments)

python manage_comments.py list --block BLOCK_ID [--limit N] [--all] [--json]
python manage_comments.py create --page PAGE_ID --text "Comment text" [--json]
python manage_comments.py create --discussion THREAD_ID --text "Reply text" [--json]
python manage_comments.py get --comment COMMENT_ID [--json]

Subcommand: list | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --block / --page | Yes | - | Block or page ID | | --limit | No | 50 | Max comments | | --all | No | False | Fetch all pages | | --json | No | False | Output raw JSON |

Subcommand: create | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --page | No* | - | Page ID for new comment | | --discussion | No* | - | Discussion thread ID for reply | | --text | Yes | - | Comment text | | --json | No | False | Output raw JSON |

*Either --page or --discussion is required

Subcommand: get | Argument | Required | Default | Description | |----------|----------|---------|-------------| | --comment | Yes | - | Comment ID | | --json | No | False | Output raw JSON |

Note: Requires "Insert comments" capability in integration settings.

When to Use: Use list to read comments on a page, create to add comments or reply to threads, get to retrieve specific comment details. Great for collaboration and feedback workflows.

Skill Import/Export

download_skill.py - Download skill from Notion

python download_skill.py PAGE_ID [PAGE_ID2 ...] [--output-dir PATH] [--no-backup] [--json]

| Argument | Required | Default | Description | |----------|----------|---------|-------------| | page_ids | Yes | - | One or more Notion page IDs (positional) | | --output-dir | No | 03-skills | Output directory for skill folder | | --no-backup | No | False | Don't backup existing skills | | --json | No | False | Output result as JSON |

When to Use: Use when user wants to import a skill from the Skills DB, download a skill bundle from Notion, or sync a skill from the team repository to local Nexus.

upload_skill.py - Upload skill to Notion

python upload_skill.py SKILL_PATH --team TEAM [--integration LIST] [--dry-run] [--as-new NAME] [--skip-validation] [--json]

| Argument | Required | Default | Description | |----------|----------|---------|-------------| | skill_path | Yes | - | Path to skill folder (positional) | | --team | Yes | - | Team: General, Solutions, Engineering, Sales | | --integration | No | - | Comma-separated integrations (Beam AI, Linear, etc.) | | --dry-run | No | False | Preview upload without pushing | | --as-new | No | - | Upload as new skill with different name | | --skip-validation | No | False | Skip SKILL.md validation | | --json | No | False | Output as JSON |

When to Use: Use when user wants to share a local skill to the team, export a skill to Notion, or publish a skill to the Skills DB for others to use.

Utility Module

rate_limiter.py - Rate limit handling (module, not CLI)

from rate_limiter import with_retry, RateLimiter, make_request_with_retry

# Decorator usage
@with_retry(max_retries=3, base_delay=1.0)
def my_api_call():
    return requests.get(...)

# Direct usage
limiter = RateLimiter(max_retries=3, base_delay=1.0)
response = limiter.execute(lambda: requests.get(...))

Features: Exponential backoff, jitter, respects Retry-After header, auto-retry on 429/5xx

When to Use: This is a library module, not a CLI script. Import it in Python code when making direct API calls that need rate limit protection (3 req/s limit).

Intelligent Error Detection Flow

When a Notion skill fails due to missing configuration, the AI should:

Step 1: Run Config Check with JSON Output

python 00-system/skills/notion/notion-master/scripts/check_notion_config.py --json

Step 2: Parse the ai_action Field

The JSON output includes an ai_action field that tells the AI what to do:

| ai_action | What to Do | |-----------|------------| | proceed_with_operation | Config OK, continue with the original operation | | proceed_with_warning | Partial config (can query/import but not export) | | prompt_for_api_key | Ask user: "I need your Notion API key. Get one at https://www.notion.so/my-integrations" | | create_env_file | Create .env file and ask user for API key | | run_setup_wizard | Run: python 00-system/skills/notion/notion-master/scripts/setup_notion.py |

Step 3: Help User Fix Issues

If ai_action is prompt_for_api_key:

  1. Tell user: "Notion integration needs setup. I need your API key."
  2. Show them: "Get one at: https://www.notion.so/my-integrations"
  3. Ask: "Paste your Notion API key here:"
  4. Once they provide it, write directly to .env: 
    # Edit .env file to add:
NOTION_API_KEY=secret_their_key_here
NOTION_SKILLS_DB_ID=2bc2cadf-bbbc-80be-af8a-d45dfc8dfa2e
  5. Re-run config check to verify

JSON Output Structure

{
  "status": "not_configured",
  "exit_code": 2,
  "ai_action": "prompt_for_api_key",
  "missing": [
    {"item": "NOTION_API_KEY", "required": true, "location": ".env"}
  ],
  "fix_instructions": [...],
  "env_template": "NOTION_API_KEY=secret_YOUR_API_KEY_HERE\nNOTION_SKILLS_DB_ID=...",
  "setup_wizard": "python 00-system/skills/notion/notion-master/scripts/setup_notion.py"
}

How Skills Reference This

Each skill loads shared resources only when needed (progressive disclosure):

notion-connect uses:

  • All discovery and query scripts
  • All page, block, user, and comment management scripts
  • All references (filter-syntax, property-types, block-types)

query-notion-db uses:

  • check_notion_config.py (validate before query)
  • api-reference.md (query patterns)
  • error-handling.md (troubleshooting)

import-skill-to-nexus uses:

  • check_notion_config.py (validate before import)
  • api-reference.md (file download API)
  • error-handling.md (troubleshooting)

export-skill-to-notion uses:

  • check_notion_config.py (validate before export)
  • api-reference.md (file upload API)
  • database-schema.md (field mapping)
  • error-handling.md (troubleshooting)

Usage Example

User says: "query my Projects database"

What happens:

  1. AI loads notion-connect skill (NOT notion-master)
  2. notion-connect SKILL.md says: "Run check_notion_config.py first"
  3. AI executes: python notion-master/scripts/check_notion_config.py
  4. AI executes: python notion-master/scripts/search_skill_database.py --db "Projects"
  5. If errors occur, AI loads: notion-master/references/error-handling.md

notion-master is NEVER loaded directly - it's just a resource library.

Version: 2.6 Created: 2025-12-10 Updated: 2025-12-11 Status: Production Ready

Changelog:

  • v2.6: Added "When to Use" sections to all 13 scripts for AI routing guidance
  • v2.5: Added comprehensive script argument documentation with usage examples and argument tables
  • v2.4: Added Intelligent Error Detection Flow with --json support for AI-guided setup
  • v2.3: Renamed query_database.py to search_skill_database.py for clarity