lovable-project
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script (
plugins/lovable/scripts/detect-architecture.py) to determine the project's architecture. This is a standard operation within the skill's intended purpose. - [DYNAMIC_EXECUTION]: The skill relies on the runtime execution of a local script provided within the repository to report project settings.
- [INDIRECT_PROMPT_INJECTION]: The skill reads several files that could be controlled by an external party if a repository is malicious or compromised.
- Ingestion points: Loads configuration and project context from
.lovable-agent/config.json,.lovable-agent/context.md, andAGENTS.md(documented inSKILL.md). - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified for these project files.
- Capability inventory: The skill possesses the capability to execute shell commands (via Python) and read/modify project files.
- Sanitization: While the skill explicitly instructs the agent to preserve secret safety by not recording sensitive values, it lacks specific sanitization for natural language instructions embedded within the ingested data.
Audit Metadata