sdd-slim-implement-learning
Pass
Audited by Gen Agent Trust Hub on May 7, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires the agent to run validation routines such as '定向测试' (targeted tests), 'typecheck', 'lint', and 'build' commands. These operations are essential for the software development lifecycle and are executed locally on the user's project code.
- [PROMPT_INJECTION]: The skill relies on untrusted external data in the form of
*.spec.mdfiles to drive its implementation logic, which represents an indirect prompt injection surface. - Ingestion points: Project specification files (
*.spec.md) are read to define the implementation tasks (described inimplement.md). - Boundary markers: The instructions do not define specific delimiters or guardrails to prevent instructions embedded within the spec files from overriding the agent's core behavior.
- Capability inventory: The agent has the capability to modify project files and execute shell commands for testing and building code.
- Sanitization: There is no evidence of input validation or content filtering for the data read from the specification files.
Audit Metadata