add-test-case
Pass
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to run the shell command
pnpm run buildto verify the generated code in Step 3. - [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it incorporates untrusted user input from
$ARGUMENTSdirectly into a code generation task. - Ingestion points: User input is ingested through the
$ARGUMENTSvariable inSKILL.mdto determine the test case name and logic. - Boundary markers: There are no boundary markers or 'ignore' instructions specified to delimit the user-provided input within the prompt templates.
- Capability inventory: The agent has the capability to write to local filesystem files (e.g.,
test-cases.ts) and execute build commands (pnpm run build). - Sanitization: The instructions lack guidance on sanitizing or escaping the user-provided input before it is interpolated into the TypeScript code template. This creates a surface for indirect prompt injection where a malicious input could attempt to break out of the string literal and insert executable code.
Audit Metadata