The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its use of external data tools.
Ingestion points: The skill instructions explicitly call for web_search and perplexity tools to retrieve external data at runtime.
Boundary markers: There are no delimiters or instructions to the agent to disregard instructions that may be embedded in the retrieved web content.
Capability inventory: While the skill retrieves external data, it does not currently exhibit high-risk capabilities such as writing to the filesystem or executing system commands.
Sanitization: No logic is present to sanitize or filter the results returned by the web search tools.
[SAFE]: The YAML frontmatter contains a typo where brain-first: exempt is used instead of the canonical brain_first: exempt. This causes the intended exemption configuration to be ignored, meaning platform security constraints remain active for this skill.

typo-frontmatter