Getting Started

Installation

Install Last Signal v0.1.0 from PyPI, verify local system prerequisites, and configure local Ollama models.

Last updated: 2026-10-31•v0.1.0

Last Signal is distributed as a lightweight Python CLI package on PyPI. It has zero heavy dependencies, relying strictly on standard library streaming subprocesses, Typer for CLI parsing, and Rich for terminal rendering.

Prerequisites

Python 3.10 or higher (Python 3.10, 3.11, 3.12, 3.13, 3.14 supported).
Operating System: Linux, macOS, or Windows (WSL recommended).
Optional: Local Ollama runtime for local AI interpretations.

Installation from PyPI

To install the official v0.1.0 release directly from PyPI, run:

bash

$pip install last-signal==0.1.0

To install the latest general release with automatic updates:

bash

$pip install last-signal

Official PyPI Release

View package releases, hashes, and release history on the official PyPI page: https://pypi.org/project/last-signal/0.1.0/

Verifying Installation

Confirm that the lastsignal executable is available in your PATH by requesting the version and help menu:

bash

$lastsignal --help

Usage: lastsignal [OPTIONS] COMMAND [ARGS]...

Ghost in the Machine: Offline-First Digital Autopsy CLI for Failed AI-Agent Executions.

Commands:

run Supervise agent execution and stream telemetry traces.

analyze Ingest trace and evaluate deterministic failure rules.

report Generate GitHub-ready post-mortem Markdown report.

doctor Check system environment, directories, and Ollama health.

Optional Local AI Setup (Ollama)

Last Signal's deterministic core operates 100% offline without any external services. However, if you want local AI root-cause hypotheses, install Ollama and pull the lightweight default model:

bash

$# Install Ollama from ollama.com, then pull the default model:

$ollama pull llama3.2

$# Verify Ollama daemon is running:

$curl http://localhost:11434/api/tags

Running Environment Doctor

Run the built-in diagnostic suite to confirm your permissions and optional model availability:

bash

$lastsignal doctor

[SUCCESS] Python Runtime: 3.12.2 (>= 3.10 required)

[SUCCESS] Workspace Permissions: Write access confirmed for .lastsignal/ and reports/

[SUCCESS] Ollama Daemon: Reachable at http://localhost:11434

[SUCCESS] Default Model: llama3.2 is installed and ready

[SUCCESS] System ready for offline agent supervision.

Installation from PyPI

To install the official v0.1.0 release directly from PyPI, run:

bash

$pip install last-signal==0.1.0

To install the latest general release with automatic updates:

bash

$pip install last-signal

Official PyPI Release

View package releases, hashes, and release history on the official PyPI page: https://pypi.org/project/last-signal/0.1.0/

Verifying Installation

Confirm that the lastsignal executable is available in your PATH by requesting the version and help menu:

bash

$lastsignal --help

Usage: lastsignal [OPTIONS] COMMAND [ARGS]...

Ghost in the Machine: Offline-First Digital Autopsy CLI for Failed AI-Agent Executions.

Commands:

run Supervise agent execution and stream telemetry traces.

analyze Ingest trace and evaluate deterministic failure rules.

report Generate GitHub-ready post-mortem Markdown report.

doctor Check system environment, directories, and Ollama health.

Optional Local AI Setup (Ollama)

Last Signal's deterministic core operates 100% offline without any external services. However, if you want local AI root-cause hypotheses, install Ollama and pull the lightweight default model:

bash

$# Install Ollama from ollama.com, then pull the default model:

$ollama pull llama3.2

$# Verify Ollama daemon is running:

$curl http://localhost:11434/api/tags

Running Environment Doctor

Run the built-in diagnostic suite to confirm your permissions and optional model availability:

bash

$lastsignal doctor

[SUCCESS] Python Runtime: 3.12.2 (>= 3.10 required)

[SUCCESS] Workspace Permissions: Write access confirmed for .lastsignal/ and reports/

[SUCCESS] Ollama Daemon: Reachable at http://localhost:11434

[SUCCESS] Default Model: llama3.2 is installed and ready

[SUCCESS] System ready for offline agent supervision.