meta-forest-model-plot
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied CSV files, creating a potential surface for indirect prompt injection if the data contains hidden instructions.
- Ingestion points: The scripts
scripts/forest_survival.pyandscripts/forest_survival.Rread a CSV file path provided via the command line as instructed inSKILL.md. - Boundary markers:
SKILL.mdincludes validation instructions and task boundaries, but does not define explicit delimiters (e.g., XML tags) for the CSV data content itself. - Capability inventory: The skill possesses the ability to write files to the local system (PNG plots and CSV summary tables) and execute bundled Python/R scripts.
- Sanitization: The implementation scripts perform data type conversion and validate that hazard ratios and confidence intervals are positive numeric values, providing a basic level of protection against non-numeric payloads.
- [SAFE]: The skill follows standard data analysis workflows. It utilizes well-known, legitimate scientific libraries (such as pandas, matplotlib, and metafor) and restricts its operations to local script execution with deterministic inputs. The included
meta-forest-model-plot_audit_result_v2.jsonappears to be a development artifact and does not contain malicious content. No network operations, credential access, or obfuscation techniques were detected.
Audit Metadata