Official $REPR Links: Bags.fm · Axiom · Jupiter · Photon
Official CA: 5WsMLk8Zb8PWTXoHev7Ry6QDHNC2KSpY5x8R13GGBAGS (on Solana)

Repr CLI

The developer context layer.

Your git history is rich with context about what you build, how you think, and how you grow. repr captures that context and makes it available everywhere — to AI agents, to applications, to you.

Use it however you need: Interview prep, performance reviews, social content, AI agent context — these are all lenses into the same underlying data. Local-first, privacy-focused, works offline.

Real Developers, Real Results

"I used repr to prep for my Meta interview in 30 minutes. Turned 2 years of commits into 8 STAR-format stories. Nailed every behavioral question."
— Sarah, Senior Backend Engineer

"Our sprint demos went from chaos to polished in 5 minutes. Just run repr commits --days 14 and export. Stakeholders love it."
— Marcus, Engineering Manager

"I run repr in a fully air-gapped environment. Zero network calls, 100% local. It's the only tool I trust for this."
— Alex, Defense Contractor

Lenses (Ways to Use Your Context)

• 🤖 AI Agent Context — MCP server lets Claude Code, Cursor, and other agents know your work history
• 🎯 Interview Prep — Generate STAR-format stories in 30 minutes (interview lens)
• 📊 Performance Reviews — Turn 6 months of work into quantified impact (resume lens)
• 🚀 Sprint Demos — Professional changelogs for stakeholders (changelog lens)
• 📱 Build in Public — Social posts from your actual work (content lens)
• 🔒 Proof of Work — Verified credentials from real commits (proof lens)
• 💼 Engineering Managers — Team summaries and sprint recaps

Why Repr

Context That Compounds

The longer you use repr, the richer your context becomes. By the time you need a resume, interview prep, or content — you have months of structured history ready to use. No other tool builds this persistent layer.

AI-Native

repr exposes your context via MCP (Model Context Protocol), so AI agents like Claude Code, Cursor, and Clawdbot can know your work history and patterns. Your coding assistant finally understands you.

# Start MCP server for AI agents
repr mcp serve

Privacy First (Not an Afterthought)

• ✅ Local-first by default — Your repos, diffs, and stories stay on your machine in ~/.repr/
• ✅ Air-gapped ready — Works in fully offline environments (defense, healthcare, finance approved)
• ✅ Bring your own model — Use local LLMs (Ollama) or your own API keys (OpenAI/Anthropic)
• ✅ Privacy audit — See exactly what data (if any) left your machine with repr privacy audit
• ✅ OS keychain — API keys never touch config files, stored in system keychain
• ✅ Zero telemetry — No tracking, no analytics, no silent uploads

Story Engine (New in v0.2.16)

repr now synthesizes Stories from your commits — coherent narratives that capture WHY and HOW you built something, not just WHAT changed.

• Generate: Create stories from your commits with repr generate
• Dashboard: Explore your work in a premium dark-mode UI with repr dashboard (or rp dashboard)
• MCP Integration: AI agents can answer questions about your implementation details

Read the Story Engine Documentation →

Time Savings

vs. Alternatives

vs. Manual brag documents:
❌ Requires discipline to maintain
❌ Easy to forget to update
❌ No structure or templates
✅ repr: Automatic, retroactive, professional templates

vs. GitHub commit history:
❌ Raw commits are cryptic
❌ No narrative or context
❌ Not interview/resume ready
✅ repr: LLM transforms commits into narratives

vs. Trying to remember at review time:
❌ Forget 80% of your work
❌ Can't quantify impact
❌ Miss your best stories
✅ repr: Never forget, always quantified

Install

macOS / Linux (Homebrew)

brew tap repr-app/tap
brew install repr

Direct Download

Grab pre-built binaries for macOS, Linux, and Windows from the latest release.

Python (pipx)

pipx install repr-cli

Quickstart (60 seconds)

# 1) Scan your repos and set up local config
repr init ~/code

# 2) Generate stories from your recent work (local LLM)
repr generate --local

# 3) See what you created
repr stories
repr story view <id>

Common workflows

For full step-by-step guides, see the documentation. Below are the quick happy-path snippets.

First-time setup

repr init ~/code
repr commits --days 7
repr generate --local

Full guide →

Daily workflow

repr hooks install --all
repr generate --local
repr review

Full guide →

Weekly reflection

repr commits --days 7
repr generate --local
repr story edit <id>
repr story feature <id>

Full guide →

Interview prep (STAR stories)

repr generate --template interview --local
repr stories
repr story view <id>

Full guide →

Generate from a specific timeframe

# Last 30 days
repr generate --days 30 --local

# Since a specific date
repr generate --since 2024-01-01 --local

# Natural language dates
repr generate --since "2 weeks ago" --local
repr generate --since monday --local

Note: repr generate automatically skips commits that have already been processed into stories. You can safely run it multiple times without creating duplicates.

Publish your profile (optional)

repr login
repr push --dry-run
repr push --all
repr profile link

Full guide →

Privacy-focused (local only)

repr privacy lock-local
repr llm configure
repr llm test
repr generate --local

Full guide →

Multi-device sync

repr login
repr sync

Full guide →

Troubleshooting

repr status
repr mode
repr doctor

Full guide →

Configure your models

Your config lives at ~/.repr/config.json.

Local LLM (Ollama/LocalAI)

repr llm configure

# or set it manually:
repr config set llm.local_api_url http://localhost:11434/v1
repr config set llm.local_model llama3.2

Bring your own API keys (BYOK)

repr llm add openai
repr llm add anthropic
repr llm use byok:openai

Privacy modes

Command help

For the full flag reference:

repr --help
repr <command> --help

Enterprise & Compliance

Air-Gapped Environments

repr works in fully offline, air-gapped environments:

# 1. Install repr (transfer binary via USB)
# 2. Install Ollama and download models offline
# 3. Lock to local-only permanently
repr privacy lock-local --permanent

# 4. Generate stories (zero network calls)
repr generate --local

Use cases:

• Defense contractors (classified environments)
• Healthcare (HIPAA compliance)
• Finance (SOX/PCI requirements)
• Stealth startups (pre-launch confidentiality)

Privacy Audit Trail

See exactly what data left your machine:

repr privacy audit --days 30

Output:

Network Activity Audit (Last 30 days)

No network activity detected.

Local operations:
  • 143 commits analyzed
  • 23 stories generated  
  • 0 cloud syncs
  • 0 API calls to repr.dev

Mode: LOCAL_ONLY (locked)

Perfect for security audits and compliance reviews.

BYOK (Bring Your Own Key)

Use your own API keys, stored securely:

# Keys stored in OS keychain (never in config files)
repr llm add openai
repr llm add anthropic

# Calls go directly to your provider, not repr.dev
repr generate  # Uses your OpenAI key

• ✅ Keys stored in macOS Keychain / Windows Credential Manager / Linux Secret Service
• ✅ repr.dev never sees your keys or data
• ✅ Full control over costs and models

Documentation

• Full Documentation — Comprehensive guides
• USER_WORKFLOWS.md — 10 detailed workflow examples
• Privacy Model — How data is protected

License

MIT License — see LICENSE.

🚀 Ready to unlock your git history?

brew install repr
repr init ~/code
repr generate --local