The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources via get_query_run_results and get_terminal_results in SKILL.md. Boundary markers are absent. Capabilities include SQL execution via run_sql_query. Sanitization is absent, relying instead on manual validation instructions to mitigate potential data poisoning.
[DYNAMIC_EXECUTION]: The skill generates SQL queries for the run_sql_query tool based on user input, which is the core analytical function of the skill.
[SAFE]: The skill uses expected vendor tools and follows best practices for data reproducibility. No external downloads from untrusted sources or malicious persistence mechanisms were identified.

allium-investigation