The Agent Skills Directory

[NO_CODE]: The skill consists entirely of Markdown instructions and pattern templates. There are no scripts, binaries, or executable files included in the skill package.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data (academic papers, articles, and transcripts), which is a known attack surface for indirect prompt injection. However, since the skill has no active capabilities such as network access or file system modification, the risk is negligible.
Ingestion points: User-supplied research content as described in SKILL.md.
Boundary markers: The reference templates in the references/ directory do not utilize specific delimiters or instructions to ignore embedded commands.
Capability inventory: No tools or dangerous system capabilities are defined or requested.
Sanitization: No sanitization or filtering of the input content is performed.
[SAFE]: No signs of obfuscation, credential theft, persistence mechanisms, or unauthorized privilege escalation were detected. The skill functions solely as a text-processing assistant.

research-intelligence