kubesense-spl
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill describes a read-only log query interface with specific security constraints, including the rejection of filesystem and network functions within the query language. It emphasizes using a validation tool (
validate-spl) to check queries before execution. - [INDIRECT_PROMPT_INJECTION]: The skill creates an ingestion surface for untrusted data by providing access to log records via the
execute-spltool. - Ingestion points: Untrusted content from logs enters the agent's context through the output of the
execute-spltool. - Boundary markers: The skill documentation does not define specific delimiters or instructions to prevent the agent from following directives potentially embedded in log messages.
- Capability inventory: The skill provides metadata discovery (
get-trace-or-log-fields), query validation (validate-spl), and query execution (execute-spl). - Sanitization: No specialized sanitization or filtering of the log
bodyis described before retrieval.
Audit Metadata