configuring-block-policy
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides MATLAB code examples for the agent to interact with policy configuration files.
- Evidence includes the use of
library.BlockPolicy.save,library.BlockPolicy.loadRaw, andlibrary.BlockPolicy.validatewithin the project environment. - [INDIRECT_PROMPT_INJECTION]: The skill processes project-level configuration data from
.satk/block-policy.json, which represents a potential attack surface for indirect prompt injection if the file content is malicious. - Ingestion points: Reads and summarizes existing
.satk/block-policy.jsonand.satk/reuse-libraries.jsonfiles. - Boundary markers: None identified; instructions do not explicitly mandate delimiters or escaping when summarizing file content for the user.
- Capability inventory: The skill has file writing capabilities through the
library.BlockPolicy.savefunction and can lookup information in the Knowledge Graph (KG). - Sanitization: The skill includes a guardrail requiring the use of
library.BlockPolicy.validate(policyData)before saving any changes.
Audit Metadata