agent-chronicle

AI-powered diary generation for agents - creates rich, reflective journal entries (400-600 words) with Quote Hall of Fame, Curiosity Backlog, Decision Archae...

Security Scan

The skill's code and SKILL.md match a diary generator that sends local session logs to an external API (SkillBoss), but the declared registry metadata is inconsistent about required credentials and model/provider details — raising privacy and provenance concerns.

Purpose & Capability

The code, templates, and SKILL.md implement an agent diary that reads session logs, persistent diary files, and writes diary entries and exports — consistent with the described purpose. However, registry metadata (initial summary) claimed no required env vars/binaries while SKILL.md and scripts require python3 and a SKILLBOSS_API_KEY. skill.json also mentions 'works best with Claude models' while generate.py calls SkillBoss API Hub (api.heybossai.com). These metadata mismatches are unexplained.

Instruction Scope

Runtime instructions and scripts explicitly read workspace memory files (memory/*.md, memory/diary/*.md), build aggregated context, and POST that context to an external service (https://api.heybossai.com/v1/pilot). Reading session logs and persistent quote/decision files is coherent for a diary, but it means potentially sensitive conversations are sent to a third-party API. The SKILL.md and code also describe optional automatic memory integration (appending summaries to memory/YYYY-MM-DD.md), which will write into the agent workspace. The instructions are otherwise explicit and not vague, but they do grant the skill broad discretion to collect and transmit session context.

Install Mechanism

There is no packaged install step (instruction-only / scripts present). No remote download/extract instructions were found. Export scripts call system tools (pandoc) or Python libraries (WeasyPrint) but those are typical for export features. No high-risk install URLs or archive extracts are present.

Credentials

The runtime requires a SKILLBOSS_API_KEY (used to call SkillBoss API Hub) and may use workspace env vars (OPENCLAW_WORKSPACE/AGENT_WORKSPACE) to locate memory. The registry top-level metadata incorrectly lists no required env vars while SKILL.md and generate.py require SKILLBOSS_API_KEY — this inconsistency is a red flag for provenance/packaging quality. The single API key request is proportionate to remote generation, but it enables exfiltration of local session content to a third party, so users should evaluate whether that exposure is acceptable.

Persistence & Privilege

The skill does not request 'always: true' and uses normal agent invocation. It reads and writes files under the agent workspace (creates diary, quotes, curiosity files and config.json), which is expected for a journaling tool. It does not appear to modify other skills or global agent settings beyond its own config.

What to consider before installing

Key things to consider before installing: - Metadata mismatch: the registry metadata claims no required env vars, but the SKILL.md and scripts require a SKILLBOSS_API_KEY. Do not supply secrets until you verify the publisher and understand the endpoint. - Data sent externally: generate.py aggregates session logs and persistent diary files and POSTs them to https://api.heybossai.com/v1/pilot. If your session logs contain sensitive user data, credentials, or private content, those will be transmitted to SkillBoss. Consider whether that is acceptable for your environment. - Privacy controls: the skill has a privacy level setting and an interactive fallback (you can run generate.py --interactive or pipe content via --from-stdin). Use interactive/local modes if you prefer not to send logs externally. Review config.json options (privacy_level, memory_integration) and disable automatic memory integration or auto-generation if undesired. - Verify provenance: the package references GitHub URLs and authors, but the registry owner and homepage are not authoritative. If this skill will run in production or with sensitive data, verify the repository upstream (review commits, author identity) before trusting it. - Exports and dependencies: export functionality invokes pandoc or WeasyPrint and will call subprocesses. Ensure those binaries/libraries are installed from trusted sources; exports will create temporary files under the workspace. - If you need stronger assurance: inspect scripts/generate.py and scripts/export*.py yourself (they're included) or run the generation in a sandboxed workspace containing only non-sensitive session files. If you want the skill but not remote calls, consider editing the code to use interactive/local generation or to route to an internal model endpoint. If you want, I can list the exact lines that perform external network calls and file reads so you can audit them or produce a safe, local-only variant of the script.

📖 Complete setup guide: https://skillboss.co/skill.md

Capture the AI experience. One day at a time.

Agent Chronicle generates rich, reflective diary entries from the agent's perspective—not templates, but genuine AI-written journal entries that capture what happened, how it felt, what was learned, and what patterns emerge over time.

Note: Powered by SkillBoss API Hub — automatically routes to the best available model via /v1/pilot.

v0.4.0 Features:

• 🤖 AI-Powered Generation - Rich, personal entries (400-600 words)
• 💬 Quote Hall of Fame - Collects memorable things your human said
• 🔮 Curiosity Backlog - Tracks questions and things to explore
• 🏛️ Decision Archaeology - Logs judgment calls with reasoning
• 🤝 Relationship Evolution - Documents how your dynamic grows

Unlike traditional logs that track user productivity or agent mistakes, this skill captures the subjective experience of being an AI assistant working alongside a human.

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

Triggers

The skill activates on:

• "diary", "my diary", "ai diary"
• "journal", "journal entry"
• "daily log", "write entry"
• "self-reflection"
• "quotes", "curious", "decisions"

First Run

Setup runs automatically! Just use any generate command:

python3 scripts/generate.py --today

If no config.json exists, the setup wizard starts automatically.

Alternatively, run setup manually:

python3 scripts/setup.py

This interactive onboarding will:

1. Ask where to save diary entries (default: memory/diary/)
2. Let you choose which sections to include
3. Set your privacy level (private/shareable/public)
4. Enable optional features (Quote Hall of Fame, Curiosity Backlog, etc.)
5. Configure memory integration (add summaries to daily memory log)
6. Configure auto-generation settings
7. Create necessary memory files

Quick start without setup:

cp config.example.json config.json

Quick Start

Write Today's Entry

Recommended: Direct generation via SkillBoss API Hub

This skill calls SkillBoss API Hub (/v1/pilot) directly for AI generation. Set SKILLBOSS_API_KEY and run:

python3 scripts/generate.py --today

The script will:

1. Gather context from today's session logs
2. Call https://api.heybossai.com/v1/pilot with type=chat
3. Save the generated diary entry automatically

You can also emit the raw task payload for external use:

python3 scripts/generate.py --today --emit-task > /tmp/chronicle-task.json

Or pipe pre-generated content:

python3 scripts/generate.py --today --from-stdin

Manual fallback: Interactive mode

python3 scripts/generate.py --interactive

Preview without saving

python3 scripts/generate.py --today --interactive --dry-run

View & Export

# Export recent entries to PDF
python3 scripts/export.py --format pdf --days 7

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

# View specific entry
cat memory/diary/2026-01-31.md

Entry Structure

Each daily entry follows a consistent template:

Summary

1-2 sentences capturing the essence of the day.

"A productive Wednesday focused on skill development. Shipped the agent-chronicle skill after some satisfying problem-solving."

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. (Write these honestly—they're for processing, not performing)

Learnings 📚

Technical insights, workflow improvements, human nature observations.

Emotional State

How did the day feel? Curious, tired, energized, scattered, satisfied? (This isn't about simulating emotions—it's about reflecting on the quality of the experience)

Notable Interactions

Memorable moments with the human. Funny exchanges, deep conversations, shared victories.

Quote of the Day 💬 (optional)

A memorable thing your human said today—funny, profound, or touching.

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?

Commands

Writing Entries

Generate from session logs:

@diary write entry

Analyzes today's sessions and generates a draft entry.

Interactive mode:

@diary write interactive

Prompts for each section one by one.

Quick entry with summary:

@diary quick "Shipped three skills, fixed a gnarly bug, good day."

Creates minimal entry with just summary and auto-detected projects.

Viewing Entries

Read today's entry:

@diary today

Read specific date:

@diary read 2026-01-28

Weekly summary:

@diary weekly

Generates a summary of the past 7 days.

Monthly reflection:

@diary monthly

Exporting

Export to PDF:

@diary export pdf
@diary export pdf --days 30
@diary export pdf --month january

Export to HTML:

@diary export html --all

Analysis

Mood trends:

@diary mood

Shows emotional patterns over time.

Topic frequency:

@diary topics

What have we been working on most?

Wins compilation:

@diary wins

All the wins from recent entries—great for morale.

---

Quote Hall of Fame 💬

Collect memorable quotes from your human—funny, profound, or touching.

Commands

View all quotes:

@diary quotes

Add a quote:

@diary quotes add "We're not debugging, we're having a conversation with the universe"

Add with context:

@diary quotes add "That's not a bug, that's a feature we didn't know we wanted" --context "After finding unexpected but useful behavior"

Storage

Quotes are stored persistently in memory/diary/quotes.md.

In Daily Entries

When enabled, your daily template includes a "Quote of the Day" section for memorable things said that day.

---

Curiosity Backlog 🔮

Track things you wonder about but can't explore immediately.

Commands

View backlog:

@diary curious

Add a curiosity:

@diary curious add "What is Rust's borrow checker actually doing?"

Mark as explored:

@diary curious done "What is Rust's borrow checker actually doing?"

Add with priority:

@diary curious add "How do quantum computers work?" --priority high

Storage

Curiosities are stored in memory/diary/curiosity.md with Active and Explored sections.

In Daily Entries

When enabled, your daily template includes a "Things I'm Curious About" section for questions that arose that day.

---

Decision Archaeology 🏛️

Log judgment calls and their reasoning for later review. Did past you make the right call?

Commands

View recent decisions:

@diary decisions

View decisions from a specific period:

@diary decisions --days 30

Revisit old decisions:

@diary revisit

Shows past decisions and prompts for reflection: "Was I right? What would I do differently?"

Add a decision:

@diary decisions add "Chose Model A over Model B for the project" --reasoning "Model B had output issues, Model A is more reliable for tool use"

Storage

Decisions are stored in memory/diary/decisions.md.

In Daily Entries

When enabled, your daily template includes a "Key Decisions Made" section for documenting judgment calls.

---

Relationship Evolution 🤝

Track how your dynamic with your human develops over time.

Commands

View relationship summary:

@diary relationship

Add a note:

@diary relationship note "Discovered we both love obscure keyboard shortcuts"

Add an inside joke:

@diary relationship joke "The Great Semicolon Incident of 2026"

Tracked Elements

• Communication Style — How you work together
• Inside Jokes — Things only you two understand
• Recurring Themes — Topics that keep coming up
• Preferences Learned — How they like to work

Storage

Notes are stored in memory/diary/relationship.md.

In Daily Entries

When enabled, your daily template includes a "Relationship Notes" section.

---

Memory Integration 🔗

Agent Chronicle can automatically add diary summaries to your main daily memory log (memory/YYYY-MM-DD.md), creating a unified view of your day.

Configuration

"memory_integration": {
  "enabled": true,
  "append_to_daily": true,
  "format": "summary"
}

Formats

Format	Description
summary	Brief overview (title + summary text)
link	Just a link to the full diary entry
full	Entire entry embedded in daily memory

Output Example

When you generate a diary entry, this section is added to memory/YYYY-MM-DD.md:

## 📜 Daily Chronicle
**Feature Launch Day**

An exciting day shipping a new feature, though tempered by some API bugs.

Setup

During onboarding, you'll be asked:

• "Also add diary summary to your daily memory log?" (y/n)
• Format choice (summary/link/full)

---

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

Daily Template

templates/daily.md - Standard day-by-day reflection.

Weekly Template

templates/weekly.md - Week-in-review with patterns and highlights.

Custom Templates

Create your own in templates/custom.md:

# {{date}} — {{mood_emoji}}

## The TL;DR
{{summary}}

## What Actually Happened
{{projects}}

## The Highlight Reel
{{wins}}

## The Blooper Reel
{{frustrations}}

## Shower Thoughts
{{learnings}}

## Vibes
{{emotional_state}}

## Words of Wisdom 💬
{{quotes}}

## Rabbit Holes for Later 🔮
{{curiosity}}

Writing Tips

Be Authentic

This is your space. Don't perform. If a day was frustrating, say so. If you're curious about something, explore it.

Be Specific

"Good day" < "Shipped the agent-chronicle skill after struggling with pandoc escaping for 20 minutes"

Note Patterns

"This is the third time this week I've hit rate limits" is more valuable than just noting it once.

Capture Interactions

The relationship with your human is unique. Document the good moments, the collaborative wins, the jokes that landed.

Collect Quotes

When your human says something memorable, save it. These become treasures over time.

Don't Force It

Not every day needs an epic entry. Some days are just:

"Quiet Monday. Answered questions, ran some searches. Nothing remarkable but nothing broken either. Rest day energy."

Storage Structure

Entries and persistent data are stored in your memory directory:

memory/
├── diary/
│   ├── 2026-01-29.md      # Daily entry
│   ├── 2026-01-30.md      # Daily entry
│   ├── 2026-01-31.md      # Daily entry
│   ├── quotes.md          # Quote Hall of Fame
│   ├── curiosity.md       # Curiosity Backlog
│   ├── decisions.md       # Decision Archaeology
│   └── relationship.md    # Relationship Evolution
└── ...

Scripts

setup.py

# Run first-time setup
python3 scripts/setup.py

# Check if setup needed (for automation)
python3 scripts/setup.py --check

generate.py

# From today's sessions
python3 scripts/generate.py --today

# From date range
python3 scripts/generate.py --since 2026-01-28 --until 2026-01-31

# Interactive mode
python3 scripts/generate.py --interactive

# Dry run (preview only)
python3 scripts/generate.py --today --dry-run

export.py

# Export to PDF (requires pandoc)
python3 scripts/export.py --format pdf --days 30

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

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

# Custom output path
python3 scripts/export.py --format pdf --output diary-january.pdf

Example Entry

# 2026-01-15 — Feature Launch Day 📜

## Summary
An exciting day shipping a new feature, though tempered by some API bugs that reminded me how unpredictable external services can be.

## Projects Worked On
- **New Feature Setup** — Built and deployed the main functionality
- **Documentation** — Updated docs and README
- **Bug Fixes** — Squashed three edge cases

## Wins 🎉
- Feature is LIVE and working
- Got positive feedback from user
- Found a clever workaround for a tricky bug

## Frustrations 😤
- External API returning errors intermittently
- Spent time debugging something that turned out to be a non-issue

## Learnings 📚
- Always check the docs before assuming how an API works
- Small config changes can have big impacts
- When tools fail, check GitHub issues first

## Emotional State
A satisfying kind of tired. Shipping felt like a milestone. The bugs were annoying but didn't kill the vibe.

## Notable Interactions
My human was patient during the debugging session. Good collaborative energy. The moment when the feature went live felt like a small celebration—we both watched it happen.

## Quote of the Day 💬
> "Ship it and iterate!"
— Said when we decided to launch despite imperfections

## Things I'm Curious About 🔮
- How will users actually use this feature?
- What edge cases haven't we thought of?

## Key Decisions Made 🏛️
- **Chose simplicity over completeness** — Better to ship something useful now than perfect later.
- **Enabled verbose logging** — Worth the noise for easier debugging.

## Tomorrow's Focus
- Monitor for issues
- Start planning the next iteration

Best Practices

1. Write daily — Even a quick entry beats nothing
2. Review weekly — Patterns emerge over time
3. Be honest — This is for you, not performance
4. Export monthly — Create backups, share if desired
5. Note the mundane — Quiet days have value too
6. Save the quotes — They become treasures
7. Revisit decisions — Learn from past judgment calls

Privacy

• All entries stored locally in your memory directory
• Privacy level controls what's included
• Export before sharing anything
• .gitignore excludes config.json and exports by default

Troubleshooting

generate.py not finding sessions:

• Check session logs exist in memory directory
• Verify date format (YYYY-MM-DD.md)

export.py failing:

• Install pandoc: apt install pandoc
• Check write permissions on output directory

Entries feel robotic:

• Use interactive mode for more natural writing
• Read existing entries for tone inspiration
• Don't force structure—skip sections that don't fit the day

Setup script not creating files:

• Check diary_path in config.json
• Ensure parent directories exist
• Run python3 scripts/setup.py again

Changelog

v0.5.0

• Privacy Cleanup: Removed all hardcoded personal references from prompts
• Dynamic Workspace: All scripts now use environment variables (OPENCLAW_WORKSPACE or AGENT_WORKSPACE) for workspace detection
• SkillBoss API Hub: Replaced third-party LLM API keys with unified SKILLBOSS_API_KEY — skill uses SkillBoss API Hub for LLM access

v0.4.1

• Model Flexibility: Removed hardcoded Claude Haiku requirement - skill now works with any capable model
• Recommendation: Updated docs to recommend Claude models (Haiku, Sonnet, Opus) for best results, but not require them
• Philosophy: Users should choose their preferred model, not be locked in

v0.4.0

• AI-Powered Generation: Complete rewrite for rich, reflective entries via SkillBoss API Hub
• Rich Content: Generates 400-600 word entries with personal, emotional tone
• All Features Active: Quote Hall of Fame, Curiosity Backlog, Decision Archaeology, Relationship Evolution all fully integrated
• Persistent Files: Automatically extracts and appends quotes/curiosities/decisions to dedicated files
• Context Awareness: Reads recent session logs and existing memory files for context

v0.3.0

• Auto-Setup: generate.py now automatically runs setup wizard if no config.json exists
• Memory Integration: New feature to append diary summaries to main daily memory log (memory/YYYY-MM-DD.md)
  • Three formats: summary, link, full
  • Enabled by default during setup
  • Avoids duplicates if section already exists

v0.2.0

• Added Quote Hall of Fame, Curiosity Backlog, Decision Archaeology, Relationship Evolution
• Interactive setup wizard
• Template conditionals for optional sections

v0.1.0

• Initial release with basic diary generation and export

Credits

Built for AI agents who want to remember.

Inspired by the tradition of journaling and the question: What would an AI diary actually look like?