The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use the glab and docker CLI tools to exchange credentials and authenticate with GitLab Artifact Registries.
[DATA_EXFILTRATION]: The skill describes writing short-lived authentication tokens to standard configuration files such as ~/.m2/settings.xml, ~/.npmrc, ~/.gradle/gradle.properties, and ~/.yarnrc.yml. These operations are performed in the context of the official GitLab CLI and include defensive measures such as using set +x and unset to prevent credential leakage in logs, and warnings against committing these files to version control.
[INDIRECT_PROMPT_INJECTION]: The skill creates a surface for incorporating user-provided hostnames and registry URLs into command execution pipelines. 1. Ingestion points: CLI arguments such as --hostname and --registry defined in SKILL.md and references/commands.md. 2. Boundary markers: None provided within the skill instructions; relies on CLI argument parsing. 3. Capability inventory: Shell execution of glab and docker for file writing, network requests, and authentication. 4. Sanitization: The skill does not perform its own validation of external hostnames or registry paths before interpolation.

glab-artifact-registry