The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill uses the Python pickle module for serialization and deserialization of internal data structures.
Evidence: The script scripts/run_deseq2_analysis.py uses pickle.dump() to save the DeseqDataSet object, and references/workflow_guide.md provides examples of using pickle.load() to retrieve saved count matrices and metadata.
Context: The pickle module is insecure against erroneous or maliciously constructed data and can be used to execute arbitrary code during the deserialization process.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-provided data files, which presents a surface for indirect prompt injection attacks.
Ingestion points: Data is loaded from external files via pd.read_csv() in both SKILL.md and scripts/run_deseq2_analysis.py.
Boundary markers: There are no specific boundary markers or instructions to the agent to ignore potentially malicious instructions embedded in the count matrices or metadata.
Capability inventory: The skill allows for local file writing (via to_csv and pickle.dump) and execution of a local Python script for analysis.
Sanitization: While the run_deseq2_analysis.py script validates that counts are non-negative integers, it does not sanitize metadata strings for potential injection patterns.

pydeseq2