math-help
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were identified in the skill. The content is strictly instructional documentation for using pre-installed mathematical tools via a CLI interface.
- [INDIRECT_PROMPT_INJECTION]: The skill documentation defines an interface for processing mathematical expressions.
- Ingestion points: User-provided mathematical strings are ingested from queries and passed to CLI scripts like
sympy_compute.pyandz3_solve.py. - Boundary markers: Examples in the documentation demonstrate the use of double quotes to encapsulate these mathematical expressions.
- Capability inventory: The tools support a wide range of symbolic and numerical computations, as well as file system operations such as generating image and HTML plots.
- Sanitization: Input sanitization and validation are expected to be managed by the underlying execution scripts, which are external to this skill file.
Audit Metadata