computer-use
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: Indirect Prompt Injection surface detected. The skill ingests data from local application windows, which may include untrusted content from web pages or external sources.
- Ingestion points: Accessibility tree data is retrieved via
alera computer get-app-stateandalera computer list-apps(SKILL.md). - Boundary markers: The instructions do not specify prompt delimiters or specific warnings to ignore embedded instructions within the ingested data.
- Capability inventory: The skill has the ability to interact with the environment using
click,set-value, andperform-secondary-action(SKILL.md). - Sanitization: The tool implements redaction for concealed fields and performs identity checks on UI elements before execution.
- [COMMAND_EXECUTION]: The skill uses shell commands via the
aleraCLI tool to inspect and operate desktop application windows. This is the primary intended functionality of the skill and is documented with specific safety boundaries and confirmation requirements.
Audit Metadata