ask-log 1.0.0a1

An AI log analyzer with chat interface

Ask Log

Ask Log is a CLI-first assistant that helps you explore, summarize, and reason about application logs through a conversational interface. Point it at a log file, ask questions, and iterate quickly.

This project uses LangChain under the hood and supports multiple LLM providers (OpenAI, Anthropic, Google), configurable via a simple guided setup. Conversations are optionally saved so you can resume later.

---

Quick links

---

Features

• Conversational log analysis over any text log
• Local vector index (FAISS) for retrieval-augmented answers
• Provider-agnostic via LangChain: OpenAI, Anthropic, Google Gemini
• Persistent conversation history per-log for continuity
• Colorful, ergonomic CLI using rich and click

---

Installation (local, editable)

No PyPI release yet. Install locally in editable mode for fast iteration.

1) Clone

git clone https://github.com/vandan-savla/ask-log.git
cd ask-log

2) Create and activate a virtual environment

• Windows (PowerShell)

py -m venv .venv
.\.venv\Scripts\Activate.ps1
python -m pip install -U pip setuptools wheel

• macOS/Linux (bash)

python3 -m venv .venv
source .venv/bin/activate
python -m pip install -U pip setuptools wheel

3) Install the project in editable mode

Core installation:

pip install -e .

---

Quickstart

Once installed, the ask-log command is available.

# 1) Configure your preferred provider and model
ask-log configure

# 2) Check configuration
ask-log status

# 3) Analyze a log file interactively; optionally save the conversation
ask-log chat --log-file /path/to/app.log --save ~/.ask-log/last-session.json

During configuration, you’ll be prompted for provider credentials and model. Supported providers are:

• OpenAI: OPENAI_API_KEY
• Anthropic: ANTHROPIC_API_KEY
• Google: GOOGLE_API_KEY

---

Commands

# Guided config flow (choose provider, model, and options)
ask-log configure

# Show current provider configuration
ask-log status

# Start an interactive session on a specific log file
ask-log chat --log-file /path/to/logfile.log --save path/to/convo.json

# Reset configuration (removes ~/.ask-log/config.yaml)
ask-log reset

Configuration is stored at ~/.ask-log/config.yaml.

---

Tips

• Use natural language questions like “What errors do you see?”, “Summarize main events”, “Any anomalies around 10:32?”
• Use --save to capture the conversation so you can resume context later.
• The first run on a large log may build a local vector index; subsequent runs will be faster.

---

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines, a suggested workflow, and development tips.

---

License

This project is licensed under the MIT License.