Openclaw Executive Assistant Webinars

OpenClaw Executive Assistant Webinars

Skill by ara.so — Hermes Skills collection.

Overview

This project provides a workshop framework for building local-only executive assistant workflows using OpenClaw. It demonstrates three core patterns:

1. Data intake review — Turn unknown files into trustworthy intake reports
2. Operational memory — Convert work residue into daily and weekly momentum docs
3. Offline communications triage — Transform exported emails into action lists

All workflows use local folders, produce reviewable markdown artifacts, and require no live integrations.

Installation

Clone the repository:

git clone https://github.com/dandenney/webinars-build-your-own-executive-assistant-with-openclaw.git
cd webinars-build-your-own-executive-assistant-with-openclaw

No dependencies required — this is a file-based workflow using OpenClaw prompts with copy/paste execution.

Repository Structure

code-along/
├── INDEX.md
├── 01-data-intake-review/
│   ├── incoming/          # Files to inspect
│   ├── prompts/
│   │   └── intake-review.md
│   ├── outputs/           # Generated reports
│   └── expected/
│       └── report-outline.md
├── 02-operational-memory/
│   ├── inbox/             # Work notes and residue
│   ├── prompts/
│   │   ├── daily-log.md
│   │   └── weekly-hype.md
│   ├── outputs/
│   ├── schedule/
│   │   ├── cron-examples.md
│   │   └── heartbeat-note.md
├── 03-offline-communications-triage/
│   ├── eml/               # Exported email files
│   ├── prompts/
│   │   └── email-triage.md
│   ├── outputs/
│   └── expected/
│       └── report-outline.md
└── mission-control/       # Optional dashboard

Workflow Patterns

1. Data Intake Review

Process unknown files in a folder and generate a structured intake report.

Setup:

cd code-along/01-data-intake-review

Workflow:

1. Place files to review in incoming/
2. Open prompts/intake-review.md to see the prompt template
3. Copy the prompt and provide context about files in incoming/
4. Save the generated output to outputs/intake-review.md

Expected output structure:

• File inventory
• Risk assessment
• Action recommendations
• Priority rankings

Example prompt pattern:

Review all files in the incoming/ directory and create an intake report that includes:
- List of all files with size and type
- Security/privacy concerns
- Recommended actions for each file
- Overall priority assessment

Format as markdown with clear sections.

2. Operational Memory

Convert scattered work notes into structured daily logs and weekly summaries.

Setup:

cd code-along/02-operational-memory

Daily log workflow:

# Daily Log Prompt Pattern
Review all notes in inbox/ from today and create a daily log with:
- Key accomplishments
- Decisions made
- Blockers encountered
- Tomorrow's priorities

Output to: outputs/daily-log.md

Weekly summary workflow:

# Weekly Hype Prompt Pattern
Review all daily logs from this week and create a weekly summary with:
- Week's highlights
- Momentum indicators
- Patterns observed
- Next week's focus areas

Output to: outputs/weekly-hype.md

Automation example (cron):

# Daily log generation at 5pm
0 17 * * * /path/to/generate-daily-log.sh

# Weekly summary on Friday at 4pm
0 16 * * 5 /path/to/generate-weekly-hype.sh

3. Offline Communications Triage

Process exported email files (.eml) and generate action-oriented triage reports.

Setup:

cd code-along/03-offline-communications-triage

Workflow:

1. Export emails as .eml files to eml/ directory
2. Use the triage prompt from prompts/email-triage.md
3. Generate report to outputs/email-triage.md

Example triage prompt pattern:

Process all .eml files in eml/ and create a triage report with:

## Urgent Actions
- Emails requiring immediate response
- Deadlines within 24 hours

## This Week
- Items needing response this week
- Grouped by topic/project

## FYI / Archive
- Informational items
- No action needed

## Delegatable
- Items that could be handled by others

For each item include:
- Sender
- Subject line
- Key points
- Recommended action

Expected output:

• Categorized action items
• Priority rankings
• Response drafts for urgent items
• Delegation opportunities

Common Patterns

File-Based Prompt Execution

All exercises follow this pattern:

# 1. Navigate to exercise directory
cd code-along/01-data-intake-review

# 2. Review the prompt template
cat prompts/intake-review.md

# 3. Provide context files
# (Files already in incoming/ or inbox/ directories)

# 4. Copy prompt + execute with OpenClaw
# (Manual copy/paste to AI assistant)

# 5. Save output
# Save response to outputs/[exercise-name].md

Markdown Output Structure

All outputs should be markdown files with:

# Report Title

**Generated:** YYYY-MM-DD HH:MM

## Executive Summary
[2-3 sentence overview]

## [Section 1]
[Detailed content]

## [Section 2]
[Detailed content]

## Recommendations
- [ ] Action item 1
- [ ] Action item 2

## Next Steps
1. Step one
2. Step two

Scheduling Automated Runs

For operational memory workflows:

# Create a simple shell script wrapper
#!/bin/bash
# generate-daily-log.sh

WORKSPACE="/path/to/code-along/02-operational-memory"
cd "$WORKSPACE"

# Your OpenClaw execution command here
# This could be an API call, CLI command, etc.
openclaw execute \
  --prompt "$(cat prompts/daily-log.md)" \
  --context "inbox/" \
  --output "outputs/daily-log-$(date +%Y-%m-%d).md"

Configuration

Environment Variables

If integrating with OpenClaw APIs:

export OPENCLAW_API_KEY=your_api_key_here
export OPENCLAW_MODEL=claude-3-5-sonnet
export WORKSPACE_ROOT=/path/to/code-along

Directory Conventions

Maintain this structure for each workflow:

• incoming/ or inbox/ — Input files
• prompts/ — Reusable prompt templates
• outputs/ — Generated reports (gitignored if desired)
• expected/ — Example outputs for reference

Troubleshooting

No outputs generated:

• Verify files exist in the input directory
• Check prompt file is readable
• Ensure output directory has write permissions

Reports missing sections:

• Review the prompt template for completeness
• Check that all input files were accessible
• Verify markdown formatting in output

Automation not running:

• Test cron syntax with crontab -l
• Check script permissions: chmod +x generate-*.sh
• Verify paths are absolute in cron jobs
• Check cron logs: grep CRON /var/log/syslog

File encoding issues:

• Ensure .eml files are UTF-8 encoded
• Check for special characters in filenames
• Verify file extensions match expected types

Best Practices

1. Review before committing — All outputs are markdown for human review
2. Version control prompts — Keep prompt templates in git
3. Iterate on prompts — Refine based on output quality
4. Archive outputs — Date-stamp reports for historical reference
5. Local-first — No API calls until you're ready to automate

Workshop Usage

To follow the webinar:

# Open the walkthrough
open webinar-runbook.html

# Work through exercises in order
cd code-along/01-data-intake-review
# ... complete exercise 1

cd ../02-operational-memory
# ... complete exercise 2

cd ../03-offline-communications-triage
# ... complete exercise 3

Each exercise builds on patterns from the previous one, demonstrating progressively sophisticated file-based AI workflows.