aiconfig-projects
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
- [DATA_EXFILTRATION] (MEDIUM): The skill is instructed in
SKILL.mdto automatically detect API keys by reading local configuration files, specifically~/.claude/config.json, and scanning environment variables. Accessing local credential stores is a sensitive operation that involves reading files outside the immediate project scope. - [COMMAND_EXECUTION] (MEDIUM): The skill provides implementation patterns in
references/env-config.mdandreferences/python-setup.mdthat involve writing to local files (such as.env,config.yaml, andk8s-secret.yaml) and executing shell commands (likecurl) to verify setup or interact with cloud secret managers. These operations modify the local development environment. - [EXTERNAL_DOWNLOADS] (LOW): The skill interacts with the LaunchDarkly API (
app.launchdarkly.com) and suggests the installation of standard Python packages (requests,python-dotenv) to facilitate project management. While expected for its primary purpose, network operations with sensitive tokens require trust.
Audit Metadata