How do I install the Digital Fte Orchestrator skill?

Run npx skillmds add majiayu000/digital-fte-orchestrator-2 in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Digital Fte Orchestrator safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Digital Fte Orchestrator?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Digital Fte Orchestrator free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Digital Fte Orchestrator?

majiayu000 (@majiayu000) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Digital Fte Orchestrator by majiayu000

Name: Digital Fte Orchestrator
Rating: 4.2 (5 reviews)
Author: majiayu000

Digital Fte Orchestrator

Main task processing loop using Claude Code with Ralph Wiggum pattern. Use when running the orchestrator, debugging task processing, configuring approval workflows, or understanding the task lifecycle. NOT when creating individual watchers (use specific watcher skills). NOT when debugging specific integrations (use integration-specific skills).

majiayu000 3e99011 2 files · 4.4 KB Updated Sep 9, 2026 567 repo stars

File contents

Digital FTE Orchestrator Skill

Core task processing engine with Ralph Wiggum loop pattern.

Quick Start

# Normal mode
python src/orchestrator.py

# Process once and exit
python src/orchestrator.py --once

# Dry run (no execution)
python src/orchestrator.py --dry-run

# Via scripts
python scripts/run.py

Task Flow

Needs_Action/ → [Claim] → In_Progress/ → [Process] → Done/
                                       ↓
                              Pending_Approval/ (if sensitive)
                                       ↓
                              [Human approves]
                                       ↓
                                  Approved/ → [Execute] → Done/

Work Zones

The orchestrator operates in different modes based on environment:

Zone	`FTE_ROLE`	Capabilities
Cloud	`cloud`	Read-only, draft creation, analysis
Local	`local`	Full execution, financial actions, posting

Cloud Zone Restrictions

All sends/posts forced to approval workflow
Cannot execute financial transactions
Cannot post to social media directly
Creates drafts for human review

Local Zone Powers

Direct execution of approved tasks
Financial actions (with audit logging)
Social media posting
WhatsApp message sending

Claim-by-Move Pattern

Atomic task claiming prevents race conditions:

# ✓ CORRECT - Atomic claim
src = Needs_Action / task_file
dst = In_Progress / task_file
src.rename(dst)  # Atomic on same filesystem

# ✗ WRONG - Race condition possible
if task_file in Needs_Action:  # Another agent could claim here!
    process(task_file)

Skill Routing

The orchestrator routes tasks to appropriate skills based on task type:

Task Type	Skill	Example Trigger
`email_*`	`sending-emails`	Email draft in queue
`whatsapp_*`	`watching-whatsapp`	WhatsApp message request
`facebook_*`	`posting-facebook`	Social media post
`invoice_*`	`managing-odoo`	Create invoice request
`calendar_*`	`managing-calendar`	Meeting request

Agentic Intelligence

Integrated with src/intelligence/agentic_intelligence.py for:

Task complexity scoring
Autonomous decision making
Learning from outcomes
Pattern recognition

Configuration

Flag	Default	Description
`--max-iterations`	10	Max retries per task
`--poll-interval`	30	Seconds between polls
`--dry-run`	false	Test without execution
`--once`	false	Process once and exit

Environment:

FTE_ROLE=local|cloud  # Determines work zone

CEO Briefing Integration

Orchestrator generates briefings via src/reports/ceo_briefing.py:

Weekly activity summaries
Key metrics and KPIs
Task completion rates
Pending approvals count

Verification

Run: python scripts/verify.py

Related Skills

sending-emails - Email sending patterns
watching-whatsapp - WhatsApp monitoring
managing-calendar - Calendar management
managing-odoo - Accounting operations
posting-facebook - Social media posting
generating-ceo-briefing - Executive reports

---name: digital-fte-orchestrator-2description: Main task processing loop using Claude Code with Ralph Wiggum pattern. Use when running the orchestrator, debugging task processing, configuring approval workflows, or understanding the task lifecycle. NOT when creating individual watchers (use specific watcher skills). NOT when debugging specific integrations (use integration-specific skills).--- # Digital FTE Orchestrator Skill Core task processing engine with Ralph Wiggum loop pattern. ## Quick Start ```bash # Normal mode python src/orchestrator.py # Process once and exit python src/orchestrator.py --once # Dry run (no execution) python src/orchestrator.py --dry-run # Via scripts python scripts/run.py ``` ## Task Flow ``` Needs_Action/ → [Claim] → In_Progress/ → [Process] → Done/ ↓ Pending_Approval/ (if sensitive) ↓ [Human approves] ↓ Approved/ → [Execute] → Done/ ``` ## Work Zones The orchestrator operates in different modes based on environment: | Zone | `FTE_ROLE` | Capabilities | |------|------------|--------------| | **Cloud** | `cloud` | Read-only, draft creation, analysis | | **Local** | `local` | Full execution, financial actions, posting | ### Cloud Zone Restrictions - All sends/posts forced to approval workflow - Cannot execute financial transactions - Cannot post to social media directly - Creates drafts for human review ### Local Zone Powers - Direct execution of approved tasks - Financial actions (with audit logging) - Social media posting - WhatsApp message sending ## Claim-by-Move Pattern Atomic task claiming prevents race conditions: ```python # ✓ CORRECT - Atomic claim src = Needs_Action / task_file dst = In_Progress / task_file src.rename(dst) # Atomic on same filesystem # ✗ WRONG - Race condition possible if task_file in Needs_Action: # Another agent could claim here! process(task_file) ``` ## Skill Routing The orchestrator routes tasks to appropriate skills based on task type: | Task Type | Skill | Example Trigger | |-----------|-------|-----------------| | `email_*` | `sending-emails` | Email draft in queue | | `whatsapp_*` | `watching-whatsapp` | WhatsApp message request | | `facebook_*` | `posting-facebook` | Social media post | | `invoice_*` | `managing-odoo` | Create invoice request | | `calendar_*` | `managing-calendar` | Meeting request | ## Agentic Intelligence Integrated with `src/intelligence/agentic_intelligence.py` for: - Task complexity scoring - Autonomous decision making - Learning from outcomes - Pattern recognition ## Configuration | Flag | Default | Description | |------|---------|-------------| | `--max-iterations` | 10 | Max retries per task | | `--poll-interval` | 30 | Seconds between polls | | `--dry-run` | false | Test without execution | | `--once` | false | Process once and exit | Environment: ```bash FTE_ROLE=local|cloud # Determines work zone ``` ## CEO Briefing Integration Orchestrator generates briefings via `src/reports/ceo_briefing.py`: - Weekly activity summaries - Key metrics and KPIs - Task completion rates - Pending approvals count ## Verification Run: `python scripts/verify.py` ## Related Skills - `sending-emails` - Email sending patterns - `watching-whatsapp` - WhatsApp monitoring - `managing-calendar` - Calendar management - `managing-odoo` - Accounting operations - `posting-facebook` - Social media posting - `generating-ceo-briefing` - Executive reports

Frequently asked questions