find-job
Warn
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: MEDIUMCREDENTIALS_UNSAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill instructions allow the agent to access a local file containing LinkedIn credentials at
../secrets/linkedin/farukz.txtto manage session expiration. - [COMMAND_EXECUTION]: The skill provides instructions for the agent to execute a shell command (
npm run pdf) in the../farukdirectory to regenerate resume assets. - [DATA_EXFILTRATION]: The skill accesses sensitive local paths across directory boundaries relative to the workspace, including resume files containing personal data and the
../secrets/directory for credentials. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external LinkedIn job descriptions and screening questions to formulate application responses.
- Ingestion points: Content captured via
browser_snapshotfrom LinkedIn job listings and application modals. - Boundary markers: Absent; the instructions do not specify delimiters or warnings to ignore instructions embedded in the external job descriptions.
- Capability inventory: The agent has capabilities for browser automation, local file reading, and shell command execution.
- Sanitization: There is no mention of sanitizing or validating the text extracted from LinkedIn before it is used to generate screening answers.
Audit Metadata