The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to load and process experimental data from various formats (ROOT, HDF5, CSV, binary). This represents an attack surface where maliciously crafted data files could potentially influence the agent's logic or be interpolated into generated code.
Ingestion points: Ingests external data files in ROOT, HDF5, CSV, and binary formats as described in SKILL.md.
Boundary markers: There are no explicit instructions or delimiters provided to the agent to treat the contents of these data files as untrusted or to ignore instructions embedded within them.
Capability inventory: The skill possesses the capability to generate and execute Python and ROOT scripts, as well as save reports and plots.
Sanitization: The instructions do not specify any sanitization or validation logic to filter out potentially malicious strings or instructions from the ingested data before processing.
[DYNAMIC_EXECUTION]: The skill's primary function is to generate and run analysis scripts. This inherent behavior involves the dynamic generation of executable code based on analysis requirements.
Evidence: SKILL.md specifies that the outputs include a reproducible Python / ROOT script and the stopCondition requires that the script runs successfully.

physics-data-analysis-root