alicloud-platform-docs-api-review-test
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script located at
tests/common/compile_skill_scripts.pyas part of its compilation check. This script is a vendor-owned resource for the author 'cinience' and is part of the skill's testing infrastructure. - [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface because it reads and parses content from a target skill's
SKILL.mdto identify API endpoints for execution. - Ingestion points: Untrusted data enters the agent context when reading the documentation of another skill at
skills/platform/docs/alicloud-platform-docs-api-review/SKILL.md. - Boundary markers: There are no specified delimiters or instructions telling the agent to ignore instructions embedded within the parsed documentation.
- Capability inventory: The skill allows for local command execution (
python3) and file system writes to theoutput/directory. - Sanitization: The instructions do not define any validation or escaping of API names or parameters extracted from the target skill before they are utilized in automated calls.
Audit Metadata