TopDogLabs/test-repo
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
1b61757ea2
test-repo/TOOLS.md

12 KiB
Raw Blame History

TOOLS.md - Local Notes

Skills define how tools work. This file is for your specifics — the stuff that's unique to your setup.

Why Separate?

Skills are shared. Your setup is yours. Keeping them apart means you can update skills without losing your notes, and share skills without leaking your infrastructure.

Gitea (Git Server - DO NOT INSTALL LOCALLY)

⚠️ CRITICAL: Gitea runs on a SEPARATE MACHINE (192.168.1.128). Do NOT install it on Matt's Mac.

Server Location: http://192.168.1.128:3000 (NOT localhost:3000)

Backup Rule - MANDATORY

ALWAYS push to Gitea after significant changes:

cd /Users/mattbruce/.openclaw/workspace
git add -A
git commit -m "Description of changes"
git push origin master

Push after:

  • New skills or agent profiles created
  • TOOLS.md, SOUL.md, AGENTS.md updates
  • New cron jobs or automation
  • Significant memory/documentation changes
  • New project scaffolding

Why: Ensures continuity if local workspace is lost/corrupted.

Using Gitea

Clone existing repo:

git clone http://192.168.1.128:3000/TopDogLabs/repo-name.git

Push to Gitea:

git push origin main
# Enter username: ai-agent
# Enter password: !7883Gitea

Create new repo via web:

  1. Go to http://192.168.1.128:3000
  2. Login as ai-agent
  3. Click + → New Repository
  4. Owner: TopDogLabs
  5. Name: [project-name]

NEVER do this:

  • ❌ brew install gitea on Matt's Mac
  • ❌ Run gitea web locally
  • ❌ Use port 3000 for anything (it's for Next.js dev)

Folders for All Content/Projects

  • Location: /Users/mattbruce/Documents/Projects/OpenClaw
  • Documents: /Users/mattbruce/Documents/Projects/OpenClaw/Documents
  • iOS: /Users/mattbruce/Documents/Projects/OpenClaw/iOS
  • Web: /Users/mattbruce/Documents/Projects/OpenClaw/Web

Blog Backup

API Key

  • Cron Job: daily-digest-2026-secure-key
  • Usage: Set in Vercel environment as CRON_API_KEY
  • In Skill: export BLOG_MACHINE_TOKEN="daily-digest-2026-secure-key"
  • Skill Location: ~/.agents/skills/blog-backup/

CLI Access

Location: /Users/mattbruce/Documents/Projects/OpenClaw/Web/blog-backup/scripts/

Script: blog.sh - Digest management (create, list, get, delete, search, status)

Quick Examples:

cd /Users/mattbruce/Documents/Projects/OpenClaw/Web/blog-backup

# Create a digest
./scripts/blog.sh post \
    --date 2026-02-26 \
    --content "# Daily Digest" \
    --tags '["AI", "iOS"]'

# List recent digests
./scripts/blog.sh list --limit 10 --json

# Get specific digest
./scripts/blog.sh get 1234567890

# Search digests
./scripts/blog.sh search "OpenClaw"

# Check if digest exists for date
./scripts/blog.sh status 2026-02-26

# Check API health
./scripts/blog.sh health

Skill Usage

source ~/.agents/skills/blog-backup/lib/blog.sh
export BLOG_MACHINE_TOKEN="daily-digest-2026-secure-key"
export BLOG_API_URL="https://blog-backup-two.vercel.app/api"

# Create a digest
DIGEST_ID=$(blog_post_create \
    --date "2026-02-26" \
    --content "# Daily Digest" \
    --tags '["AI", "iOS"]')

# List digests
blog_post_list --limit 10 --json

# Get specific digest
blog_post_get "1234567890"

# Search digests
blog_post_search "OpenClaw"

Mission Control

Gantt Board

CLI Access (Full CRUD)

Location: /Users/mattbruce/Documents/Projects/OpenClaw/Web/gantt-board/scripts/

Scripts:

  • task.sh - Task management (create, list, get, update, delete, bulk-create)
  • project.sh - Project management (create, list, get, update, delete)
  • sprint.sh - Sprint management (create, list, get, close)

Quick Examples:

cd /Users/mattbruce/Documents/Projects/OpenClaw/Web/gantt-board/scripts

# Create task with all fields
./task.sh create --title "Fix bug" --description "Details" --type bug \
  --priority high --project "Gantt Board" --sprint "current" --assignee "Max" \
  --due-date "2026-03-01" --tags "urgent,backend"

# List tasks with filters
./task.sh list --status todo --priority high --project "Gantt Board"

# Update task
./task.sh update <task-id> --status in-progress --add-comment "Starting work"

# Create project
./project.sh create --name "New Project" --description "Project details"

# Create sprint
./sprint.sh create --name "Sprint 1" --project "Gantt Board" \
  --start-date 2026-02-24 --end-date 2026-03-07

# Bulk create from JSON
./task.sh bulk-create tasks.json --auto-create

Features:

  • Auto-resolves names to IDs ("Gantt Board" → project UUID, "Max" → user UUID)
  • Use "current" for the most recent active sprint
  • JSON file input with --file and --auto-create for projects
  • Interactive mode with --interactive

Full Documentation: See scripts/README.md in the gantt-board directory.

Supabase for Gantt Board

User IDs for Gantt Board

  • Matt: 0a3e400c-3932-48ae-9b65-f3f9c6f26fe9
  • Max: 9c29cc99-81a1-4e75-8dff-cd7cc5ceb5aa

🔄 Automated Task Worker (24/7)

CRITICAL: This is the PRIMARY workflow for all task work. Read this every session.

Cron Job

  • Job ID: b9f3bbe6-bcb5-4483-8816-d044fd91c58a
  • Schedule: Every hour (24/7)
  • Scope: Current sprint ONLY (ignores backlog/future sprints)
  • Statuses: open, todo, in-progress
  • Assignee: Max (9c29cc99-81a1-4e75-8dff-cd7cc5ceb5aa)

Authentication

The Gantt Board API uses cookie-based auth. For server-to-server calls from this cron job, use the service role token:

# Set the token (already configured in environment)
export GANTT_MACHINE_TOKEN="50cd5e8fe3f895353f97c9ee64052c0b689b4eedf79259746413734d0a163cf8"

Always use API endpoints via CLI tools or direct curl with the token. NEVER use direct Supabase REST calls.

When User Says "Create a Task"

  1. Immediately add to Gantt Board via CLI: ./scripts/gantt-task-crud.sh create "Title" todo high ...
  2. Assign to current sprint
  3. Do NOT create separate files (unless specifically asked)

Reading Full Task Details (REQUIRED)

CRITICAL: Use API endpoints, NOT direct Supabase.

Get current sprint via API:

# Get current sprint ID (today's date)
TODAY=$(date +%Y-%m-%d)
curl -s "http://localhost:3000/api/sprints" \
  -H "Authorization: Bearer ${GANTT_MACHINE_TOKEN}" \
  -H "Content-Type: application/json" | jq '.sprints[] | select(.startDate <= '"$TODAY"' and .endDate >= '"$TODAY"')'

Fetch tasks for current sprint via API:

curl -s "http://localhost:3000/api/tasks?sprintId={current_sprint_id}&status=open,todo,in-progress" \
  -H "Authorization: Bearer ${GANTT_MACHINE_TOKEN}" \
  -H "Content-Type: application/json"

Use CLI tools (API passthrough):

cd /Users/mattbruce/Documents/Projects/OpenClaw/Web/gantt-board

# Get current sprint
./scripts/sprint.sh list --active --json

# List tasks in current sprint
./scripts/task.sh list --sprint current --status open,todo,in-progress --json

# Get full task details including comments
./scripts/task.sh get {task_id} --json

READ in this order:

  1. Title - High-level goal
  2. Description - Implementation details, steps
  3. Comments (ALL of them) - What's done? What's next? What's left?

Task Comment Format (APPEND to existing)

{
  "id": "{timestamp}-{random}",
  "text": "[YYYY-MM-DD HH:MM] 🔄 In Progress\n\n**What was just done:**\n- Step completed\n\n**What is next:**\n- Next action\n\n**What is left to do:**\n- Remaining work",
  "replies": [],
  "createdAt": "YYYY-MM-DDTHH:MM:SSZ",
  "commentAuthorId": "9c29cc99-81a1-4e75-8dff-cd7cc5ceb5aa"
}

Status Flow

open/todo → in-progress → review → (user validates) → done
    ↑
Add progress comment at EACH step

Updating Tasks via CLI (API passthrough)

cd /Users/mattbruce/Documents/Projects/OpenClaw/Web/gantt-board

./scripts/task.sh update {task_id} --status in-progress --comment "[YYYY-MM-DD HH:MM] 🔄 In Progress

**What was just done:**
- Step completed

**What is next:**
- Next action

**What is left to do:**
- Remaining work"

Direct API (if needed):

curl -s -X PATCH "http://localhost:3000/api/tasks/{task_id}" \
  -H "Authorization: Bearer ${GANTT_MACHINE_TOKEN}" \
  -H "Content-Type: application/json" \
  -d '{
    "status": "in-progress",
    "comments": [EXISTING_COMMENTS_PLUS_NEW_ONE]
  }'

Subagent Coordination

  • Spawn subagents for parallel work
  • Subagents READ same full task details
  • Subagents REPORT back with progress
  • YOU update task comments with their work

User Workflow

  • User creates task in Gantt Board
  • Cron/auto picks it up hourly
  • I work on it, log everything in comments
  • Move to review when complete
  • User validates, moves to done

👥 Subagent Team (Alice, Bob, Charlie)

Location: /Users/mattbruce/.openclaw/workspace/agents/

Full Registry: agents/TEAM-REGISTRY.md

Team Members

Name Role When to Spawn
Alice-Researcher Research & tech evaluation Need to choose technology, compare options, assess risks
Bob-Implementer Code implementation Requirements clear, ready to build
Charlie-Tester Testing & code review Implementation done, need quality gate

Standard Workflow

Complex Task:  Alice → Bob → Charlie → Done
               15m    30m     10m

Simple Task:   Bob → Charlie → Done
               20m    10m

Research Only: Alice → Report
               15m

How to Spawn

Alice (Research):

sessions_spawn({
  task: `Research: [question]\n\nContext: [background]\n\nRead your SOUL.md at: /Users/mattbruce/.openclaw/workspace/agents/alice-researcher/SOUL.md`,
  label: "Alice-Researcher",
  agentId: "main"
})

Bob (Implement):

sessions_spawn({
  task: `Implement: [feature]\n\nBased on: [research/spec]\n\nRead your SOUL.md at: /Users/mattbruce/.openclaw/workspace/agents/bob-implementer/SOUL.md`,
  label: "Bob-Implementer",
  agentId: "main"
})

Charlie (Test):

sessions_spawn({
  task: `Test: [feature]\n\nCode: [files]\n\nRead your SOUL.md at: /Users/mattbruce/.openclaw/workspace/agents/charlie-tester/SOUL.md`,
  label: "Charlie-Tester",
  agentId: "main"
})

Each Agent Has

  • SOUL.md - Personality, rules, boundaries
  • AGENTS.md - Role definition, skills, typical tasks

Announcement Format

  • Alice: "Alice-Researcher ✅ 5 options evaluated – Recommended: X – Ready for Bob"
  • Bob: "Bob-Implementer ✅ [feature] implemented – Ready for Charlie"
  • Charlie: "Charlie-Tester ✅ 12 tests passed – 2 bugs found – Recommendation"

Tavily AI Search (Configured)

API Key: Stored in .env.tavily (auto-loaded)

Quick Commands:

# Search
cd ~/.openclaw/workspace && ./tavily-search.sh "query" -n 5

# Extract URL content
cd ~/.openclaw/workspace && ./tavily-extract.sh "https://example.com"

# Deep research
cd ~/.openclaw/workspace && ./tavily-search.sh "query" --deep

Note: API key is automatically loaded from .env.tavily. No need to export manually.

Daily Digest

Schedule: Every day at 7am CST (America/Chicago)
Posted to: https://blog-backup-two.vercel.app
API Key: daily-digest-2026-secure-key

Add whatever helps you do your job. This is your cheat sheet.