The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run to execute local simulator binaries (ngspice, ltspice, and xyce) and the pdftotext utility for data extraction. These executions are core to the skill's function, are limited to specific local binaries, and use sanitized input labels for file generation.
[EXTERNAL_DOWNLOADS]: The skill performs network requests to well-known electronics distributors, including DigiKey, Mouser, element14, and LCSC, via urllib.request. These requests are used to retrieve technical component parameters (like gain-bandwidth product or quiescent current) from official domains and are consistent with the skill's primary engineering purpose.
[INDIRECT_PROMPT_INJECTION]: The skill has an indirect attack surface as it processes external data from schematic analysis JSON files and PDF datasheets to parameterize simulations.
Ingestion points: Reads schematic analysis output (analysis.json) in scripts/simulate_subcircuits.py and parses PDF content in scripts/spice_spec_fetcher.py.
Boundary markers: Standard JSON and PDF parsing is used; however, no specific 'ignore instruction' markers are applied to the processed component values.
Capability inventory: The skill has the ability to execute shell commands and perform network operations.
Sanitization: The skill implements sanitization for net names (_sanitize_net) and filesystem labels (_make_label) to ensure that data ingested from schematics does not lead to command injection or path traversal.

spice