The Agent Skills Directory

[PROMPT_INJECTION]: No override instructions, role-play injections, or attempts to bypass safety filters were found. The workflow is strictly focused on scientific data analysis.
[DATA_EXFILTRATION]: No sensitive file access or unauthorized network operations were identified. The listed tools relate to public bioinformatics databases (GWAS Catalog, Open Targets, ChEMBL, FDA).
[CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or credentials were found in the file. Guidance for environment configuration is absent but implied for the mentioned tools.
[REMOTE_CODE_EXECUTION]: The skill does not contain any patterns for downloading and executing remote scripts or commands.
[COMMAND_EXECUTION]: No shell command execution or subprocess spawning patterns were detected. The skill outlines a logical workflow using defined tool interfaces.
[EXTERNAL_DOWNLOADS]: The skill references reputable and well-known bioinformatics services (EBI GWAS Catalog, Open Targets, ChEMBL, FDA) for data retrieval, which is consistent with its stated purpose.
[OBFUSCATION]: The content is presented in clear, human-readable Markdown with no encoded strings, hidden characters, or homoglyph substitutions.
[DYNAMIC_EXECUTION]: There is no evidence of runtime code generation, compilation, or unsafe deserialization.
[INDIRECT_PROMPT_INJECTION]: While the skill processes external trait names and bioinformatics data, it lacks capabilities for arbitrary file writes or network exfiltration of sensitive data, presenting a minimal attack surface.

tooluniverse-gwas-drug-discovery