The Agent Skills Directory

[SAFE]: The skill provides standard instructional content and templates for the Opentrons Protocol API. All code examples and scripts demonstrate legitimate usage of the platform's features for liquid handling and hardware module control.
[EXTERNAL_DOWNLOADS]: The skill references the opentrons Python package. This is the official and necessary library for controlling Opentrons hardware and is a trusted dependency in this context.
[INDIRECT_PROMPT_INJECTION]: The skill documents the protocol.bundled_data property, which allows protocols to ingest external data files (e.g., CSV). While this is an ingestion surface for untrusted data, the skill does not demonstrate or encourage any unsafe operations (such as eval or command execution) on the ingested content.
Ingestion points: references/api_reference.md (via protocol.bundled_data).
Boundary markers: Not explicitly used in basic templates.
Capability inventory: Controls lab hardware (pipettes, modules) through the official opentrons API.
Sanitization: No specific data sanitization logic is implemented in the provided templates.

alterlab-opentrons