The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on shell commands (curl, bash) and Python scripts to interact with the Moonraker API and process printer data. These commands are executed on the host system to perform network operations and file management.
[INDIRECT_PROMPT_INJECTION]: The skill fetches printer.cfg and related configuration files from a remote, user-provided IP address. These external files are ingested into the agent's context and could potentially contain malicious instructions meant to override agent behavior.
Ingestion points: Configuration files like printer.cfg and mainsail.cfg are fetched via curl and parsed using regex (SKILL.md Phase 1, Step 2).
Boundary markers: None. There are no explicit instructions for the AI to treat the contents of the fetched configuration files as untrusted data.
Capability inventory: The skill has the ability to execute network requests (curl), run scripts (bash), and write to the local file system (C:/Users/<user>/Documents/printer_fix.cfg).
Sanitization: The skill uses regex-based extraction in diagnose.sh to isolate values, which provides some structural validation but does not sanitize the natural language context.
[DATA_EXFILTRATION]: The skill reads and transmits printer configuration data to the agent's context for processing. It also saves configuration changes to a local path (C:/Users/<user>/Documents/printer_fix.cfg) and uploads them to the printer via a POST request to the Moonraker API.

klipper-manager