The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of shell commands via the uv tool to manage dependencies and perform protocol simulations (e.g., opentrons_simulate). These operations are intrinsic to the skill's purpose of scientific protocol validation.
[INDIRECT_PROMPT_INJECTION]: The skill defines an attack surface for indirect prompt injection through the use of CSV-based runtime parameters.
Ingestion points: The documentation (e.g., references/api_reference.md) describes the add_csv_file method, which allows the protocol to ingest external tabular data provided by an operator.
Boundary markers: The provided protocol templates do not include explicit delimiter boundaries or instruction-ignore blocks for CSV data.
Capability inventory: The agent has access to the Bash, Write, and Edit tools, allowing for file modification and command execution based on the logic influenced by the ingested data.
Sanitization: The references/protocol_authoring.md file correctly identifies this risk and instructs authors to manually validate and parse all CSV content before issuing commands, though this is not enforced programmatically in the templates.

opentrons-integration