The Agent Skills Directory

Prompt Injection (SAFE): No evidence of instructions intended to bypass safety filters or override agent behavior.
Data Exposure & Exfiltration (SAFE): No hardcoded secrets, sensitive system paths (e.g., .ssh, .aws), or unauthorized network exfiltration patterns were identified.
Obfuscation (SAFE): Content is clear and readable with no use of Base64, zero-width characters, or homoglyphs.
Unverifiable Dependencies (SAFE): No external package installations or remote script executions are present in the skill.
Indirect Prompt Injection (LOW): The skill possesses an ingestion surface for untrusted data with associated high-privilege capabilities.
Ingestion points: Processes local source code via Read, Glob, and Grep, and web content via mcp__chrome-devtools__*.
Boundary markers: Absent; the instructions do not explicitly tell the agent to ignore instructions embedded within the code or web pages it analyzes.
Capability inventory: Includes Bash, Write, and Edit, which could be exploited if the agent obeys instructions found in processed data.
Sanitization: No sanitization or validation logic is defined for the content being read.

testing