dawnlog

A CLI tool that generates daily developer standup reports by reading your git commits across multiple repos, asking for your plan for today, and using an LLM to fill in a markdown template.

Installation

npm install -g dawnlog

Requires Node.js 20+.

From source

git clone https://github.com/yyalim/dawnlog.git
cd dawnlog
npm install
npm run build
npm link

Quick Start

dawnlog

On first run, the setup wizard will guide you through:

1. Adding your git repo paths
2. Choosing an LLM provider (Ollama, Anthropic, or OpenAI)
3. Setting your API key (if using a cloud provider)
4. Configuring output location

Then every morning:

dawnlog
# Type your plan for today, press Ctrl+D
# Your standup report is saved to ~/dawnlogs/dawnlog-YYYY-MM-DD.md

Commands

dawnlog                            # Generate today's standup (interactive)
dawnlog --today "Review PRs"       # Skip the interactive prompt
dawnlog --provider ollama          # Override provider for this run
dawnlog --dry-run                  # Print prompts without calling the LLM or saving a file
dawnlog --since 2025-01-22         # Query commits from a specific date (e.g. after a bank holiday)
dawnlog --since 2025-01-22 --dry-run  # Combine flags freely

dawnlog config                     # Show config help
dawnlog config --show              # Print current configuration as JSON
dawnlog config --edit              # Re-run the full setup wizard
dawnlog config --set <key>=<value> # Set a single config value
dawnlog config --add-repo <path>   # Add a repo to the list
dawnlog config --remove-repo <path># Remove a repo from the list

dawnlog provider                   # Interactively switch the default LLM provider

Configuration

Config is stored at ~/.dawnlog/config.json. You can edit it directly or use dawnlog config.

{
  "repos": ["/absolute/path/to/repo1", "/absolute/path/to/repo2"],
  "llm": {
    "provider": "ollama",
    "model": "gemma4"
  },
  "outputDir": "~/dawnlogs",
  "templatePath": "/path/to/templates/standup.md",
  "author": "[email protected]",
  "ticketBaseUrl": "https://yourco.atlassian.net/browse"
}

| Field | Description | Default | |-------|-------------|---------| | repos | Absolute paths to git repositories to scan | [] | | llm.provider | ollama, anthropic, or openai | ollama | | llm.model | Model to use (provider-specific) | Provider default | | llm.apiKey | API key (falls back to env var) | — | | llm.baseUrl | Custom base URL (Ollama or OpenAI-compatible) | — | | outputDir | Directory where reports are saved | ~/dawnlogs | | templatePath | Path to the output markdown template | Bundled templates/standup.md | | systemPromptPath | Path to the LLM system prompt template | Bundled templates/system-prompt.md | | author | Filter git commits by author name or email | All authors | | ticketBaseUrl | Base URL for ticket ID linkification | Disabled |

Editing config values

# Switch provider and model
dawnlog config --set llm.provider=ollama
dawnlog config --set llm.model=gemma4

# Update API key
dawnlog config --set llm.apiKey=sk-ant-...

# Set Ollama base URL (if not running on default port)
dawnlog config --set llm.baseUrl=http://localhost:11434

# Manage repos
dawnlog config --add-repo ~/projects/my-api
dawnlog config --remove-repo ~/old-project

# Other fields
dawnlog config --set author="Jane Doe"
dawnlog config --set ticketBaseUrl=https://myco.atlassian.net/browse
dawnlog config --set outputDir=~/Documents/standups
dawnlog config --set systemPromptPath=~/my-system-prompt.md

Settable keys: llm.provider, llm.model, llm.apiKey, llm.baseUrl, outputDir, templatePath, systemPromptPath, author, ticketBaseUrl

API Key Environment Variables

API keys can be set via environment variables instead of storing them in the config file:

• ANTHROPIC_API_KEY for Anthropic
• OPENAI_API_KEY for OpenAI

LLM Providers

Ollama (default)

Runs models locally — no API key or internet connection needed.

Setup:

# Install Ollama from https://ollama.com, then:
ollama serve
ollama pull gemma4   # default — or mistral, llama3.1:8b, qwen2.5:7b, etc.

Default model: gemma4. Default base URL: http://localhost:11434. Override with --set llm.baseUrl=... if needed.

Anthropic

Uses Claude models. Set ANTHROPIC_API_KEY or add it via the setup wizard. Default model: claude-haiku-4-5.

dawnlog config --set llm.provider=anthropic
dawnlog config --set llm.model=claude-haiku-4-5

OpenAI

Uses GPT models. Set OPENAI_API_KEY or add it via the setup wizard. Supports a custom baseUrl for OpenAI-compatible APIs (Groq, Together, etc.).

dawnlog config --set llm.provider=openai
dawnlog config --set llm.model=gpt-4o

Adding a Custom Provider

1. Create src/llm/myprovider.ts implementing the LLMProvider interface
2. Add a case "myprovider" to src/llm/index.ts
3. Add "myprovider" to the union type in src/config.ts

Customizing Templates

There are two template files you can customize independently:

Output template (templates/standup.md)

Defines the structure of the generated report. Copy and edit it, then point dawnlog at it:

cp /path/to/dawnlog/templates/standup.md ~/my-standup.md
# Edit ~/my-standup.md to your liking
dawnlog config --set templatePath=~/my-standup.md

The LLM fills in the placeholders — you can add, remove, or rename sections freely.

Placeholders:

| Placeholder | Description | |-------------|-------------| | {{YESTERDAY_DATE}} | Last working day's date (e.g. "Thu 02 Apr 2026") | | {{TODAY_DATE}} | Today's date | | {{YESTERDAY_SUMMARY}} | Synthesized summary of yesterday's commits, grouped by repo | | {{TODAY_PLAN}} | Your plan for today | | {{BLOCKERS}} | Blockers, or "None" | | {{WORKLOAD}} | Inferred workload: Full / Half day / Morning only / etc. |

System prompt (templates/system-prompt.md)

Controls the instructions sent to the LLM. Useful for tuning output style or improving results with a specific local model:

cp /path/to/dawnlog/templates/system-prompt.md ~/my-system-prompt.md
# Edit ~/my-system-prompt.md to your liking
dawnlog config --set systemPromptPath=~/my-system-prompt.md

The output template is passed to the LLM via the user message at runtime, so the system prompt is focused purely on instructions and examples.

Output

Reports are saved to ~/dawnlogs/dawnlog-YYYY-MM-DD.md. Example:

---
Yesterday (Thu 02 Apr 2026):

[my-api]
BWD-8682 — RBAC & permissions system
Backend:
- Defined 17 permissions and 4 roles with resolvePermissionsFromRoles
- Added requirePermissions middleware for payment admin routes
- Created GET /v1/internal/me/permissions endpoint

[dashboard]
BWD-9100 — Dashboard scaffold & dark mode
Frontend:
- Scaffolded Next.js app with Tailwind and shadcn/ui
- Added transactions table with pagination and date range filter
- Added dark mode toggle to dashboard settings

Today (Fri 03 Apr 2026):

Review open PRs and address feedback on BWD-8682

Blockers: None
Workload: Full
---

Weekend & bank holidays

If you're generating a report after a long weekend or bank holiday, use --since to specify the start date manually:

dawnlog --since 2025-01-20   # picks up commits from Jan 20 onwards

Development

npm run dev         # Run with tsx (no build needed)
npm run build       # Compile TypeScript → dist/
npm run lint        # Type-check only
npm test            # Run tests with Vitest
npm run test:watch  # Watch mode

Planned Features

• [ ] dawnlog post — post output to Slack via webhook
• [ ] dawnlog week — weekly summary across all logs in outputDir
• [ ] npx dawnlog support (publish to npm)

License

MIT