The Agent Skills Directory

[SAFE]: The skill contains no executable code, remote script downloads, or obfuscated content. Its primary function is to provide structured instructions for the AI agent to evaluate manuscripts against the journal's publicly known criteria.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest data from external sources and local files.
Ingestion points: The Official-submission checklist directs the agent to read local files (../../resources/source-basis.md, ../../resources/official-source-map.md) and search live websites for author instructions.
Boundary markers: None explicitly defined to separate external data from system instructions.
Capability inventory: The skill is limited to information retrieval and text generation; it lacks capabilities for subprocess execution, file writing, or network exfiltration.
Sanitization: None present. While this creates a theoretical ingestion surface, the risk is negligible as the agent's capabilities are restricted to generating submission advice based on the retrieved information.

nature-communications