The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of local shell execution to perform Puppet operations. Evidence includes calls to puppet, curl, jq, find, and sed across various management tasks.
[DATA_EXFILTRATION]: The discovery script accesses sensitive Puppet SSL files, specifically private keys located in /etc/puppetlabs/puppet/ssl/private_keys/, to authenticate requests to the Puppet Server status API. While this is a standard requirement for Puppet administration, it represents an access pattern to highly sensitive credentials.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting untrusted data that is then interpolated into shell commands and API queries.
Ingestion points: Variables such as $NODE, $MODULE, and $ENVIRONMENT are accepted as arguments in multiple scripts within SKILL.md.
Boundary markers: None are present to delimit user-provided data from command logic.
Capability inventory: The skill has access to puppet, curl (including POST/GET to localhost and the Puppet Server), find, and sed within SKILL.md.
Sanitization: There is no evidence of input validation or sanitization for the provided variables before they are used in shell command interpolation or URL construction.

managing-puppet