workflow-write-integration-test
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted inputs, including user prompts and repository source code (handlers and entities), which creates a surface for indirect prompt injection where malicious instructions could influence the agent's behavior during code generation or testing.
- Ingestion points: The workflow takes the user's prompt as context and explicitly requires reading handler, entity, and event source files (SKILL.md).
- Boundary markers: There are no explicit instructions or delimiters defined to treat the ingested source code or user input as untrusted or to ignore embedded instructions.
- Capability inventory: The agent has the capability to write files (specs, tests, documentation), spawn sub-agents via
spawn_agent, and execute shell commands via the test runner. - Sanitization: The skill does not define any validation or sanitization steps for the data processed from the repository or user input.
- [COMMAND_EXECUTION]: The skill triggers the execution of shell commands based on configuration found in project files.
- Evidence: The
$integration-test-verifystep is instructed to run thequickRunCommanddefined indocs/project-config.json(SKILL.md).
Audit Metadata