Skip to content

A Claude skill that maintains project context in markdown files. Define scope and constraints, manage active tasks, document decisions and learnings. Everything in .chronicle/ becomes the source of truth for ongoing AI collaboration.

License

Notifications You must be signed in to change notification settings

ricflams/chronicle-skill

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Chronicle

Chronicle captures the context gems — truths, decisions, lessons — from each Claude session and tucks them into markdown files right in your repo.

This means every new conversation starts with everything you've already figured out. Claude's built-in memory eventually scrolls out of the context window while Chronicle's markdown-captured context sticks around.

What's in it for you? First, context that keeps building — Claude gets genuinely smarter about your project over time. Second, a living record of your project's journey stored in your repo: the truths you've nailed down, the decisions you've made, the lessons you've learned. It grows organically as you work.

And don't worry about bloat — Chronicle is smart about tokens. It only passes in the most relevant parts of your journey, archiving the rest as time goes by. Your context stays lean even as your project history grows.

Think of Chronicle as Claude's sidekick for context — it keeps tabs on the important project-stuff and feeds it back into your conversations, making Claude smarter about your project with every session. All just stored as markdown-files in a .chronicle folder.


Quick Start

Install:

claude skill add https://github.com/ricflams/chronicle-skill

Use:

cd ~/my-project
claude
You: "use Chronicle"

That's it. Chronicle initializes, then runs automatically — prompting you when decisions, lessons, or sessions are worth capturing.

You can invoke Chronicle in fresh projects or in existing projects. Either works just fine.


What You Get

.chronicle/
├── truth.md              # Project DNA (purpose, scope, constraints)
├── principles.md         # Enduring beliefs that guide decisions
├── decisions/            # YYYYMMDD-HHMM-slug.md (ADRs, never archived)
├── lessons/              # YYYYMMDD-HHMM-slug.md (failures & insights, 60-day retention)
├── sessions/             # YYYYMMDD-HHMM-slug.md (timeline, 14-day retention)
└── archive/
    ├── lessons/          # Auto-archived monthly
    └── sessions/         # Auto-archived weekly

Plus:

  • Protocols in CLAUDE.md (~90 lines)
  • User guide in USING-CHRONICLE.md
  • All files git-tracked and human-readable

What Chronicle Does

Decision trail - ADRs with rationale preserved
Lessons learned - Structured failure documentation
Session narrative - Timeline of project evolution
Project truth - Purpose, scope, constraints, risks

What Chronicle Does NOT Do

Task management → Use Claude Tasks
AI context preservation → Use Claude Memory
Replace git commits → Complements with rationale


Use Cases

Team Projects

  • Shared decision history visible to all
  • Onboarding new members
  • Enriched project context from your coworkers

Long-Running Projects

  • Git-tracked evolution over months/years
  • Decision archaeology ("why did we choose X?")
  • Learning from documented failures

Solo Projects (when useful)

  • Future you needs context after 6 months
  • Want git-trackable decision history
  • Planning to hand off eventually

How It Works

Chronicle runs automatically once initialized:

  • Decisions captured when significant choices are made
  • Lessons logged when failures or surprises occur
  • Principles added when patterns emerge across decisions
  • Sessions archived weekly, lessons monthly (deterministic, conflict-free)
  • Everything git-tracked and human-readable

Philosophy

Chronicle treats understanding the journey as just as valuable than the actual produced code and artifacts.

Code can be regenerated from good context. Let's keep that around.

From .chronicle/ alone, a new team member should be able to:

  1. Understand what you're building and why
  2. Know what's been tried and what worked
  3. See what principles guide decisions
  4. Continue effectively without repeating past mistakes

This is institutional memory - knowledge that outlives individual contributors and persists across team changes.


Features

Team-first - Git-tracked, human-readable, works without Claude
Structured - ADRs for decisions, lessons for failures, principles for patterns
Automatic - Calendar-based archival, no manual intervention
Conflict-free - Timestamped files (YYYYMMDD-HHMM-slug.md), deterministic rules
Token-efficient - Always-load core ~2.5K tokens, selective ~7K total
Transparent - All markdown files, directly editable


Skill Structure

chronicle-skill/
├── SKILL.md                    # Skill definition for Claude CLI
├── USING-CHRONICLE.md          # User guide (also created in projects)
└── templates/
    ├── truth-template.md       # Project DNA template
    ├── principles-template.md  # Principles format
    ├── decision-template.md    # ADR template
    ├── lesson-template.md      # Lesson format
    ├── session-template.md     # Session log format
    └── claude-template.md      # CLAUDE.md protocols

Epilogue

Documentation

Contributing

Contributions welcome! Please open an issue to discuss proposed changes.

License

MIT

Author

Richard Flamsholt [email protected]

About

A Claude skill that maintains project context in markdown files. Define scope and constraints, manage active tasks, document decisions and learnings. Everything in .chronicle/ becomes the source of truth for ongoing AI collaboration.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published