AI Configs
Overview
Section titled “Overview”Evidence: AI Configs
Description: Parse AI tool configuration files for provider, model, and permission metadata
Category: AI
Platform: linux
Short Name: aicfg
Is Parsed: Yes
Sent to Investigation Hub: Yes
Collect File(s): No
Background
Section titled “Background”AI tools store provider, model, permission, and trust settings in files such as claude.json, settings.json, and cli-config.json. Those settings determine which models are used and whether the agent can act without approval.
Data Collected
Section titled “Data Collected”This collector gathers structured data about AI tool configurations, including provider, model, permission mode, trust overrides, and secret-like value indicators.
Collection Method
Section titled “Collection Method”This collector scans per-user and project-scope config files and parses JSON content to extract provider, model, and permission metadata and to flag risky configurations.
Forensic Value
Section titled “Forensic Value”Reveals which AI providers and models are configured, whether auto-approve or manual permission modes are in use, and whether configs contain secret-like values. Risk flags highlight permission bypasses and broad tool allowlists.