bruno-api
Warn
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: MEDIUMDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [DATA_EXFILTRATION]: The skill instructs the agent to read
.envand.env.examplefiles to extract configuration parameters likebase_url. Reading these sensitive files for configuration hints is classified as a data exposure risk. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project data, including source code and build configurations, which represents a surface for indirect prompt injection.
- Ingestion points: The agent reads source files, decorators, and project manifests (e.g.,
package.json,build.gradle,requirements.txt) to extract route and auth logic. - Boundary markers: The skill incorporates a 'Confirm' phase (Phase 3) requiring explicit user approval of the inferred plan before proceeding to file generation.
- Capability inventory: The skill executes file-write operations to the project's
bruno/directory as part of Phase 4 inSKILL.md. - Sanitization: Route normalization and naming conventions (kebab-case) are applied to the generated artifacts.
- [NO_CODE]: The skill contains instructions and reference documentation but does not include any executable scripts or binaries.
Audit Metadata