token-budget
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: No malicious patterns, suspicious commands, or security risks were identified. The skill is entirely focused on performance optimization and context management.
- [NO_CODE]: The skill consists exclusively of markdown instructions and metadata. It does not contain any executable scripts, binary files, or external package dependencies.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a protocol for loading and processing external project files (e.g., progressive loading, snippet fetching). While ingesting external data is a potential attack surface, these instructions focus on efficiency rather than execution and do not introduce new vulnerabilities.
Audit Metadata