jk
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external Jenkins controllers which could contain malicious instructions.
- Ingestion points: The agent is instructed to read Jenkins logs (
jk log), job configurations (jk job config), and build artifacts (jk artifact download) as part of its normal workflow. - Boundary markers: No explicit delimiters or boundary markers are specified in the instructions to isolate data retrieved from Jenkins from the agent's control context, increasing the risk of the agent obeying instructions embedded in logs or configs.
- Capability inventory: The skill possesses extensive administrative capabilities for modifying the Jenkins environment, such as creating or configuring jobs (
jk job create,jk job configure), managing credentials (jk cred create-secret), and installing plugins (jk plugin install). - Sanitization: The instructions do not mention sanitizing or validating Jenkins output before the agent processes it.
- [COMMAND_EXECUTION]: The skill's primary purpose is to execute system commands via the
jkCLI tool to interact with Jenkins infrastructure. This includes powerful administrative operations such as managing nodes, plugins, and credentials which could be misused if the agent is compromised via prompt injection.
Audit Metadata