Skylos FAQs

Question 1

What programming languages does Skylos support?

Accepted Answer

While primarily focused on Python (using AST), Skylos also offers multi-language support for TypeScript (using Tree-sitter). It provides built-in parsers and specific rule sets for dead code, security, and quality issues across both languages.

Question 2

What is Skylos and what does it do?

Accepted Answer

Skylos is a privacy-first Python SAST (Static Application Security Testing) tool that leverages hybrid static analysis and AI agents to detect security vulnerabilities (e.g., SQLi, secrets), dead code, and code quality issues with high precision, bridging the gap between traditional analysis and AI.

Question 3

Does Skylos support local LLMs for privacy-sensitive environments?

Accepted Answer

Absolutely. Skylos is designed with privacy in mind and fully supports 100% local LLMs such as Ollama and LM Studio. This allows you to perform AI-powered analysis without your code ever leaving your local machine, ensuring data confidentiality.

Question 4

How does Skylos ensure high precision and minimize false positives?

Accepted Answer

Skylos uses a hybrid engine (AST + optional Local/Cloud LLM) to understand dynamic Python patterns, unlike traditional linters. It employs confidence scoring, LLM reasoning for verification, and an optional `--trace` mode to validate findings against actual runtime execution, significantly reducing false positives.

Question 5

Can Skylos automatically fix identified issues?

Accepted Answer

Yes, Skylos includes an Automated Remediation Agent. It can autonomously patch security flaws and remove dead code (`skylos agent fix`). Furthermore, `skylos agent remediate` provides an end-to-end solution to scan, fix, test, and even open Pull Requests.

Skylos

Skylos

Key Features

Use Cases

Key Features

Use Cases