skills/openclaw/skills/tdd-workflow/Gen Agent Trust Hub

tdd-workflow

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [Prompt Injection] (HIGH): The skill is highly vulnerable to Indirect Prompt Injection (Category 8) because it ingests untrusted user requirements and triggers automated code-generation and testing cycles.\n
  • Ingestion points: User-provided feature descriptions (e.g., "Implement [feature] with TDD") captured in SKILL.md.\n
  • Boundary markers: Absent. The skill provides no instructions to downstream tools (like /sw:tdd:cycle or tdd-orchestrator) to treat the user's feature description as data rather than instructions.\n
  • Capability inventory: The skill invokes commands and agents that perform file writes, code generation, and shell execution (e.g., /sw:tdd:red, /sw:tdd:green, and npm test).\n
  • Sanitization: Absent. The skill lacks any mechanism to filter or sanitize instructions embedded within the user's intent strings.\n- [Command Execution] (MEDIUM): The skill's documentation and example interactions (Example 1 in SKILL.md) explicitly instruct the agent to execute local build and test commands like npm test. This capability, when combined with the lack of input sanitization, provides a direct path for executing arbitrary code if a user manages to inject malicious scripts into the test suite via the TDD cycle.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 10:34 AM