daytona-cloud-server
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local shell scripts located in the
.devcontainer/directory (e.g.,test-server-on-daytona.sh,create-daytona-openwork-server-snapshot.sh) to provision and manage development environments. - [DATA_EXPOSURE]: Uses
daytona execandtailto read application logs from the server sandbox (e.g.,/tmp/den-api.log). This is standard diagnostic behavior for development workflows. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data in the form of server health responses and log file contents. While this provides a theoretical injection surface if the server output is malicious, the operations are restricted to health monitoring and log inspection within a controlled sandbox environment.
- Ingestion points: Server health endpoints (
/health) and log files (/tmp/*.log) (SKILL.md) - Boundary markers: None explicitly defined for log processing.
- Capability inventory: Shell execution (
bash), network health checks (curl), and remote log reading (daytona exec). - Sanitization: None specified for log output inspection.
Audit Metadata