# 👻 Phantom-Shell — The Haunted Terminal

Описание и применение

Phantom-Shell — локальный первый ИИ-интерполайттер для терминалов, превращающий загадочные ошибки командной строки в структурированные проверки с доказательствами. Он captures историю выполнения команд, сохраняет её в SQLite, investigates проблемы с помощью Ollama и сохраняет проверенную информацию для будущих сессий.

Задача

Developers spend a surprising amount of time reconstructing what happened before a failure. Which command caused it? Was the same issue encountered before? Does the configuration actually support the proposed explanation?

Результат

Phantom-Shell turns terminal failures into a persistent investigation trail. It captures command execution information, output, exit codes, and working-directory context into SQLite, uses Ollama for local AI investigation, and provides a human-controlled approval workflow for proposed actions. The dashboard at http://127.0.0.1:8765/ displays the investigation timeline locally.

Рабочий процесс

• Run a command through Phantom-Shell so its execution information, output, exit code, and working-directory context are captured
• Display captured events and output with the phantomshell report command
• Review investigation activity through the local dashboard at http://127.0.0.1:8765/
• Inspect project information through read-only investigation tools
• Approve or reject proposed actions rather than giving AI unrestricted control

Доказательства

The basic capture workflow has been demonstrated: phantomshell run --no-model -- bash -lc 'echo "PhantomShell capture works"' phantomshell report

Источник

The existing automated test suite passes 25 tests in 0.82 seconds. Basic one-shot command capture and event reporting have also been demonstrated. However, passing unit tests is not the same as proving every end-to-end workflow.

Источник

The dashboard runs locally, with the displayed status indicating that inference uses Ollama through a loopback connection rather than external model API calls.

Источник

One important design decision was to separate AI reasoning from action execution. Generating a plausible fix is not the same as proving that it is correct, and proposing an action should not automatically authorize it.

Источник