The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references the @evenrealities/even_hub_sdk package. This is a vendor-owned resource originating from the skill author ('even-realities') and is consistent with the skill's stated purpose of modifying plugins for that platform.
[DATA_EXFILTRATION]: No network operations or credential harvesting patterns were detected. The skill uses local file system tools (Read, Write, Edit) to process plugin code locally.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface as it reads and analyzes untrusted source code provided via $ARGUMENTS.
Ingestion points: Files identified via Glob and Read tools based on the provided path in SKILL.md.
Boundary markers: Absent; the skill directly processes file content.
Capability inventory: Includes Bash, Write, and Edit tools, allowing for file system modifications.
Sanitization: The skill follows structured heuristics to identify state variables, but does not explicitly describe sanitization of the input code. However, the risk is inherent to the developer-tooling nature of the skill and no malicious exploitation patterns are present.
[PROMPT_INJECTION]: The skill instructions do not contain any patterns attempting to override agent behavior, bypass safety filters, or extract system prompts.

background-state