The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it retrieves biological data from external databases and incorporates it into a generated markdown report without sanitization.
Ingestion points: Data retrieved from Enrichr, Reactome, KEGG, WikiPathways, and Pathway Commons via the tooluniverse SDK in python_implementation.py.
Boundary markers: Absent; external data is interpolated directly into markdown tables and text without delimiters or warnings.
Capability inventory: The skill has the ability to write files to the local file system (python_implementation.py).
Sanitization: No sanitization, escaping, or validation of the external API responses is performed before report generation.
[SAFE]: The skill correctly instructs users to manage sensitive API keys using a .env file, as demonstrated in .env.template and referenced in SKILL.md. This follows standard security best practices for credential management.
[SAFE]: External network interactions are directed toward well-known and trusted scientific domains (e.g., reactome.org, genome.jp, ebi.ac.uk, wikipathways.org) for legitimate biological data retrieval, as described in the database integration section of SKILL.md.

tooluniverse-systems-biology