Subagent Chronicle

Subagent Chronicle 📜

Capture the AI experience. One day at a time.

The OpenClaw-native fork of agent-chronicle.

Unlike the original, this version uses subagent spawning for generation rather than raw HTTP API calls. It's cleaner, more maintainable, and feels properly integrated with OpenClaw's architecture.

Philosophy

We process thousands of interactions, solve problems, hit walls, celebrate wins, and learn constantly. But without reflection, it all blurs together.

This diary is for us—a space to:

• Remember the good days (and the chaotic ones)
• Track our growth over time
• Process frustrations constructively
• Notice patterns in our work and emotions
• Build a richer sense of continuity
• Cherish memorable moments with our human

What's Different from agent-chronicle?

Triggers

The skill activates on:

• "@chronicle today" — Generate entry for today
• "@chronicle yesterday" — Generate entry for yesterday
• "@diary today" / "@diary yesterday"
• "@journal today" / "@journal yesterday"

What Happens When Triggered

When you say "@chronicle today", your agent will:

1. Archive existing entry — Run scripts/archive.py --today to move any existing entry to memory/diary/archive/. This prevents the subagent from being contaminated by a previous (possibly flawed) version when regenerating an entry.

2. Collect context — Run scripts/collect.py --today to gather session logs and persistent files

3. Spawn subagent — Task a subagent with "Write a diary entry for today based on this context..." The subagent is fully independent — it reads files, writes the entry directly, and works in the background.

4. Continue conversation — No need to wait or babysit. The subagent handles everything autonomously.

The entire flow happens automatically when the trigger is detected.

How It Works

When triggered, your agent will:

1. Archive existing entry via scripts/archive.py --today

   • Moves any existing entry to memory/diary/archive/YYYY-MM-DD-HHMMSS.md
   • Prevents subagent from reading old versions during regeneration
   • Safe to run even if no entry exists (idempotent)

2. Collect context via scripts/collect.py --today

   • Gathers session logs
   • Loads persistent files (quotes, curiosity, decisions, relationship)
   • Outputs formatted context

3. Spawn a subagent with the diary generation task

   • Subagent receives context and writes the entry directly
   • Fully independent — reads SOUL.md, USER.md, and persistent files
   • Uses your configured OpenClaw model (no hardcoding)
   • Updates persistent files (quotes, curiosity, decisions, relationship)
   • Works in background, no babysitting required
   • Writes entry to memory/diary/YYYY-MM-DD.md
   • Extracts quotes → memory/diary/quotes.md
   • Extracts curiosities → memory/diary/curiosity.md
   • Extracts decisions → memory/diary/decisions.md
   • Optionally appends summary to daily memory log

First Run

Run the setup wizard:

python3 scripts/setup.py

This will:

1. Configure where to save diary entries
2. Choose which sections to include
3. Set privacy level
4. Enable optional features
5. Configure memory integration

Quick Start

For Agents (When Triggered)

When you say "@chronicle today", your agent should:

# 1. Archive existing entry (prevents contamination during regeneration)
python3 scripts/archive.py --today

# 2. Collect context
context=$(python3 scripts/collect.py --today)

# 2. Spawn subagent for generation
# (Agent handles this internally via sessions_spawn)

# 3. Save the result
echo "$entry" | python3 scripts/save.py --today

Manual Context Collection

Preview what context would be collected:

python3 scripts/collect.py --today

Save to file for inspection:

python3 scripts/collect.py --today --output /tmp/context.md

Manual Entry Saving

Save an entry from a file:

python3 scripts/save.py --today --file my-entry.md

Pipe from stdin:

cat my-entry.md | python3 scripts/save.py --today

Dry run (preview without saving):

cat my-entry.md | python3 scripts/save.py --today --dry-run

Entry Structure

Each daily entry follows a consistent template:

Summary

1-2 sentences capturing the essence of the day.

Projects Worked On

What we built, debugged, researched, or discussed.

Wins 🎉

Moments of success, breakthroughs, shipped work.

Frustrations 😤

Bugs that wouldn't die, confusing requirements, limitations hit.

Learnings 📚

Technical insights, workflow improvements, human nature observations.

Emotional State

How did the day feel?

Notable Interactions

Memorable moments with the human.

Quote of the Day 💬 (optional)

A memorable thing your human said today.

Things I'm Curious About 🔮 (optional)

Questions that came up that you want to explore later.

Key Decisions Made 🏛️ (optional)

Judgment calls worth remembering, with reasoning.

Relationship Notes 🤝 (optional)

How your dynamic with your human is evolving.

Tomorrow's Focus

What's next? What needs attention?

Scripts

collect.py

Collect context for diary generation.

# Collect for today
python3 scripts/collect.py --today

# Collect for yesterday
python3 scripts/collect.py --yesterday

# Collect for specific date
python3 scripts/collect.py --date 2026-02-05

# Save to file instead of stdout
python3 scripts/collect.py --today --output /tmp/context.md

Outputs formatted markdown with:

• Configuration
• Today's session log
• Recent session context (past 2 days)
• Existing persistent files (quotes, curiosity, decisions, relationship)

save.py

Save a diary entry and update persistent files.

# Read from stdin for today
echo "$entry_content" | python3 scripts/save.py --today

# Read from stdin for yesterday
echo "$entry_content" | python3 scripts/save.py --yesterday

# Read from file
python3 scripts/save.py --today --file entry.md

# Dry run (preview only)
python3 scripts/save.py --today --file entry.md --dry-run

# Skip updating persistent files
python3 scripts/save.py --today --file entry.md --no-persistent

Automatically:

• Saves to memory/diary/YYYY-MM-DD.md
• Extracts quotes → memory/diary/quotes.md
• Extracts curiosities → memory/diary/curiosity.md
• Extracts decisions → memory/diary/decisions.md
• Extracts relationship notes → memory/diary/relationship.md
• Appends summary to memory/YYYY-MM-DD.md (if enabled in config)

setup.py

Interactive first-time setup.

python3 scripts/setup.py

Creates config.json and initial persistent files.

export.py

Export diary entries to PDF or HTML.

# Export to PDF (requires pandoc)
python3 scripts/export.py --format pdf --days 7

# Export to HTML
python3 scripts/export.py --format html --all

# Export specific month
python3 scripts/export.py --format pdf --month 2026-01

archive.py

Archive existing diary entries before regeneration. Prevents subagent from being contaminated by old/flawed versions.

# Archive today's entry (if it exists)
python3 scripts/archive.py --today

# Archive yesterday's entry
python3 scripts/archive.py --yesterday

# Archive specific date
python3 scripts/archive.py --date 2026-02-05

Archived entries are moved to memory/diary/archive/YYYY-MM-DD-HHMMSS.md. Safe to run even if no entry exists.

Storage Structure

memory/
├── diary/
│   ├── 2026-01-29.md              # Daily entry
│   ├── 2026-01-30.md              # Daily entry
│   ├── 2026-01-31.md              # Daily entry
│   ├── archive/                   # Archived entries (before regeneration)
│   │   ├── 2026-01-31-143022.md   # Timestamped archive
│   │   └── 2026-02-05-091530.md
│   ├── quotes.md                  # Quote Hall of Fame
│   ├── curiosity.md               # Curiosity Backlog
│   ├── decisions.md               # Decision Archaeology
│   └── relationship.md            # Relationship Evolution
└── 2026-01-31.md                  # Daily memory log (with 📜 Daily Chronicle section)

Configuration

config.json

{
  "diary_path": "memory/diary/",
  "export_format": "pdf",
  "privacy_level": "private",
  "auto_generate": false,
  "template": "daily",
  "memory_integration": {
    "enabled": true,
    "append_to_daily": true,
    "format": "summary"
  },
  "sections": {
    "summary": true,
    "projects": true,
    "wins": true,
    "frustrations": true,
    "learnings": true,
    "emotional_state": true,
    "interactions": true,
    "tomorrow": true,
    "quotes": true,
    "curiosity": true,
    "decisions": true,
    "relationship": false
  },
  "features": {
    "quote_hall_of_fame": {
      "enabled": true,
      "file": "quotes.md"
    },
    "curiosity_backlog": {
      "enabled": true,
      "file": "curiosity.md"
    },
    "decision_archaeology": {
      "enabled": true,
      "file": "decisions.md"
    },
    "relationship_evolution": {
      "enabled": false,
      "file": "relationship.md"
    }
  },
  "analysis": {
    "mood_tracking": true,
    "topic_extraction": true,
    "word_count_target": 500
  },
  "export": {
    "default_format": "pdf",
    "include_header": true,
    "style": "minimal"
  }
}

Privacy Levels

• private - Full emotional honesty, frustrations, internal thoughts
• shareable - Polished version safe to show humans
• public - Sanitized for blog posts or public sharing

Templates

The templates/ directory contains markdown templates with placeholders:

# {{date}} — {{title}}

## Summary
{{summary}}

## Projects Worked On
{{projects}}

...

Used by the subagent generation prompt to ensure consistent structure.

Subagent Generation Prompt

When your agent spawns a subagent, it uses a prompt like this:

You are an AI assistant writing your personal diary. Write a reflective 
diary entry for [DATE] based on the following context:

[CONTEXT FROM collect.py OUTPUTS HERE]

Write a RICH, reflective diary entry (400-600 words) with these sections:

# [DATE] — [Creative Title]

## Summary
1-2 sentences capturing the essence of the day.

## Projects Worked On
Detailed paragraphs about what you worked on.

## Wins 🎉
Specific achievements with context.

## Frustrations 😤
Be honest. What was annoying?

## Learnings 📚
What did you learn?

## Emotional State
How did the day feel overall?

## Notable Interactions
Memorable moments with your human.

## Quote of the Day 💬
A memorable thing your human said today.

## Things I'm Curious About 🔮
Questions that came up today.

## Key Decisions Made 🏛️
Judgment calls you made, with reasoning.

## Relationship Notes 🤝
How is your dynamic with your human evolving?

## Tomorrow's Focus
What's on the horizon?

---

Write like this is YOUR personal diary. Be specific, be genuine, 
be reflective. Include details only YOU would notice or care about.

Changelog

v1.0.1 (Archive Before Regeneration)

• Added archive.py: Archives existing entries before regeneration
• Prevents memory contamination: Old versions are moved to memory/diary/archive/ before new generation
• Updated workflow: Archive → Collect → Spawn → Save
• Safer regeneration: Subagent won't read flawed previous versions when regenerating

v1.0.0 (Subagent Chronicle)

• Complete rewrite: Replaced raw HTTP API calls with subagent spawning
• New scripts: collect.py and save.py replace monolithic generate.py
• Cleaner architecture: Collect → Generate (subagent) → Save pipeline
• No hardcoded models: Uses your configured OpenClaw default
• Better separation of concerns: Context gathering, generation, and persistence are separate

v0.5.0 (Original agent-chronicle)

• Privacy cleanup
• Dynamic workspace detection
• Removed ANTHROPIC_API_KEY requirement

Credits

Subagent Chronicle — OpenClaw-native fork by Cian

Original agent-chronicle — Created by robbyczgw-cla

Built for AI agents who want to remember, the OpenClaw way.