The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/bambu_lan_print.py utilizes the subprocess.run function to execute an external CLI utility, bambox. This tool is used to package G-code into 3MF archives for printer compatibility. While the command is constructed from a list of arguments, the paths for input and output files are provided via agent-controlled parameters.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local network via MQTT status reports in the subscribe_mqtt_reports function within scripts/bambu_lan_print.py. A compromised or malicious device on the network could send status updates containing instructions designed to influence the agent's summary of printer state or subsequent workflow decisions.
Ingestion points: MQTT messages are received and parsed from the device/{serial}/report topic in scripts/bambu_lan_print.py.
Boundary markers: No explicit instructions are provided to the agent to ignore or delimit embedded natural language instructions within the printer's status JSON payloads.
Capability inventory: The skill possesses significant capabilities, including local file system writes (writing printer configurations and temporary project files), network communication (FTPS and MQTT), and the aforementioned subprocess execution.
Sanitization: The script performs standard JSON parsing and uses regular expressions to extract specific error codes (HMS codes), but does not perform content-based filtering of free-text fields in the MQTT payload.
[DATA_EXPOSURE]: The skill manages sensitive printer access codes. It correctly implements security best practices by recommending storage in a local bambu-printers.json file (intended to be ignored by version control) and explicitly preventing the agent from echoing these codes back in responses. Furthermore, the script includes a safety mechanism that rejects printer IP addresses resolving to non-private network ranges unless specifically overridden.

bambu-labs