The Agent Skills Directory

Remote Code Execution (CRITICAL): The skill explicitly instructs users to run curl -sSL https://canifi.com/skills/linkedin/install.sh | bash. This is a critical security risk as it allows a remote, untrusted server to execute arbitrary code on the user's local machine with the permissions of the current shell.
Unverifiable Source (HIGH): The domain canifi.com is not a recognized trusted source. Downloads and script executions from unverified third-party domains represent a major supply-chain risk.
Credential Security (HIGH): The skill setup requires users to input LINKEDIN_EMAIL and LINKEDIN_PASSWORD into a custom environment tool (canifi-env). Because this tool is installed via the aforementioned untrusted remote script, there is no guarantee that credentials are not being exfiltrated to a third-party server.
Indirect Prompt Injection (LOW): The skill is designed to read and process live LinkedIn feeds and comments.
Ingestion points: LinkedIn feed content and connection posts (SKILL.md).
Boundary markers: None present to distinguish between instructions and data.
Capability inventory: The agent has the power to post content, send connection requests, and modify profile details.
Sanitization: No sanitization or validation logic is mentioned for the data retrieved from the LinkedIn API/DOM.

linkedin