agent-tdd-london-swarm
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a
posthook to runnpm test --if-presentif apackage.jsonfile is present in the workspace. This is standard, expected behavior for an agent designed for testing and development tasks. - [INDIRECT_PROMPT_INJECTION]: The agent processes source code and test definitions to perform its tasks, creating an inherent surface for indirect prompt injection if those files contain malicious instructions.
- Ingestion points: Project source code, test files, and
package.jsonconfigurations are read and processed during the TDD cycle. - Boundary markers: The skill does not explicitly define delimiters for separating instructions from the code it processes.
- Capability inventory: The skill can execute shell commands via the
npm testhook. - Sanitization: No specific sanitization or validation of the ingested code is described within the skill instructions.
Audit Metadata