The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is configured to automatically write files to the local file system. Step 4 of the mandatory workflow in SKILL.md requires the agent to use the Write tool to save style profiles to the ./output/style-profiles/ directory by default, without needing explicit user consent for each save operation.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it processes untrusted user-provided articles to perform analysis and rewriting tasks.
Ingestion points: User-provided articles are ingested during the initial collection phase (referenced in SKILL.md, Step 1).
Boundary markers: The workflow does not define or implement boundary markers or instructions to isolate the provided text as pure data, leaving the agent vulnerable to instructions embedded within the articles.
Capability inventory: The skill possesses the capability to write files to the local disk via the Write tool (referenced in SKILL.md, Step 4).
Sanitization: No input validation or sanitization is performed on the article content before it is processed or persisted to the filesystem.

lovstudio-style-clone