The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches the official Jenkins GPG key and repository configuration from pkg.jenkins.io, a well-known service. This is a standard and safe procedure for repository management.
[COMMAND_EXECUTION]: Employs sudo for high-privilege operations including package installation, service orchestration, and user permissions management. These are necessary for the skill's primary function.
[DATA_EXFILTRATION]: Accesses the sensitive file path /var/lib/jenkins/secrets/initialAdminPassword to retrieve the mandatory setup secret. While this is the standard initialization procedure, it constitutes access to a sensitive file.
[PROMPT_INJECTION]: Contains a surface for indirect prompt injection, as the agent is instructed to process and execute logic from a Jenkinsfile stored in remote SCM.
Ingestion points: Remote Git repositories via the checkout scm step in SKILL.md.
Boundary markers: No explicit delimiters or isolation instructions are present to mitigate risks from untrusted pipeline code.
Capability inventory: Capabilities include shell command execution (sh), container management (docker), and credential handling.
Sanitization: Relies on external security analysis tools but lacks internal sanitization of the pipeline script content itself.

cicd-jenkins-debian