The Agent Skills Directory

[COMMAND_EXECUTION]: Several scripts utilize subprocess.run to orchestrate the job intake workflow. Specifically, browser_preflight.py executes osascript commands to query macOS system events and verify if Google Chrome or Firefox are running. Other scripts like finalize_capture.py, apify_capture.py, and run_job_listener.py use subprocess calls to execute secondary Python scripts within the skill's own directory structure (e.g., calling run_job_listener.py or mirror_to_sqlite.py). These executions are deterministic and do not involve shell interpolation of untrusted data.
[EXTERNAL_DOWNLOADS]: The apify_capture.py script makes network requests to the official Apify API (https://api.apify.com/v2) using urllib.request.urlopen. These requests are used to trigger web scraping tasks and download job data. Apify is a well-known web scraping service, and this interaction is essential for the skill's primary function of capturing job postings from LinkedIn and Greenhouse.
[DATA_EXFILTRATION]: The skill processes job application data and tracker files (applications.md). While it sends configuration data (task IDs and actor inputs) to Apify, this is required to facilitate the scraping service. No evidence was found of sensitive local credentials or system files being exfiltrated to unauthorized third parties.

job-intake