embedded-build-workflow
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for invoking standard embedded development tools like PlatformIO, west, and ESP-IDF. These tools are necessary for the skill's primary function and the instructions emphasize obtaining authorization before execution.
- [DATA_EXFILTRATION]: The skill includes built-in protections against accidental credential exposure. The
scripts/inspect_build.pyscript automatically redacts sensitive information (usernames/passwords) from URLs found in configuration files, and the markdown instructions specifically warn against sharing raw configuration data containing secrets. - [INDIRECT_PROMPT_INJECTION]: As the skill ingests external project files, it acknowledges the potential for malicious content in build scripts. It mitigates this risk by recommending manual review of custom scripts and by returning structured argument arrays for tool execution rather than raw shell strings.
- [SAFE]: The included Python helper scripts follow secure coding practices, including restricting file access to the designated project directory (preventing path traversal) and enforcing size limits on processed inputs to prevent resource exhaustion.
Audit Metadata