The Agent Skills Directory

[SAFE]: The skill consists of extensive documentation and reference materials based on established scientific principles (GRADE, Cochrane Risk of Bias, Bradford Hill criteria). No malicious intent or suspicious behavior was detected.
[COMMAND_EXECUTION]: The skill provides instructions for the agent to execute a local Python script (scripts/generate_schematic.py) to generate scientific diagrams. This is a functional feature designed to enhance research documentation through visual aids.
[PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it is designed to process and analyze external, untrusted scientific documents while having access to sensitive tools like Bash, Write, and Edit.
Ingestion points: External scientific papers, research abstracts, and claims provided to the agent for critique (mentioned in SKILL.md and multiple reference files).
Boundary markers: None identified; the instructions do not specify the use of delimiters or 'ignore' commands when processing external text.
Capability inventory: The skill has access to shell command execution (Bash) and file modification (Write, Edit).
Sanitization: No explicit sanitization or validation of external input is defined in the instructions.
The risk is considered acceptable (LOW) given the analytical nature of the skill and the lack of any malicious override patterns in the skill's own code.

scientific-critical-thinking