The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external, untrusted data including YouTube transcripts and Google search results.
Ingestion points: Untrusted content is ingested through the InfraNodus:analyze_text and InfraNodus:analyze_google_search_results tools as specified in SKILL.md and references/graph-to-script.md.
Boundary markers: Absent. The instructions do not provide explicit delimiters or "ignore previous instructions" safety markers when processing ingested text.
Capability inventory: Low risk. No file-system write operations, sensitive file reads, or arbitrary shell execution capabilities were detected in the provided skill files.
Sanitization: Absent. There is no evidence of sanitization or filtering for the external content before it is processed by the agent.
[NO_CODE]: The skill consists entirely of Markdown instructions and YAML configuration. No executable scripts, binaries, or complex configuration files (such as package.json or requirements.txt) are present, significantly reducing the direct attack surface.
[SAFE]: All external URLs and metadata point to the official domains and repositories of the author, InfraNodus (infranodus.com and github.com/infranodus). No signs of typosquatting or malicious redirection were found.

youtube-viral-optimizer