dev-tooling-audit
Warn
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: MEDIUMDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill accesses the AI agent's internal configuration file at
~/.claude/settings.jsonto inspect themcpServersconfiguration. This file is sensitive as it contains agent settings and potentially credentials or tokens for external services. - [COMMAND_EXECUTION]: In verify mode, the skill performs functional checks by executing system binaries and developer tools (e.g.,
typescript-language-server,pytest,eslint) to confirm they are functional. - [PROMPT_INJECTION]: The skill processes project-specific data which could contain malicious instructions.
- Ingestion points: Project configuration files such as
package.json,pyproject.toml,Cargo.toml, and.vscode/settings.json. - Boundary markers: None used to delimit project content from instructions.
- Capability inventory: Shell command execution via
ls,grep,which, and functional execution of tool binaries. - Sanitization: No validation or sanitization is performed on the content of ingested files before processing.
Audit Metadata