Observations

[999AICoder]

This utility has been incredibly helpful in my job search. Below are a few observations and suggestions after extended use:

Suggestions:

• Updating the Repo Locally: Pulling updates is tricky since modified files are part of the repo. Suggestion: include <file_name>_example.yaml and prompt users to create proper files before running. This also prevents using defaults by accident.

• Inconsistent Naming: YAML parameters use both camel case and underscores. Consistency would improve configuration and maintenance.

• Company Blacklist: Not working on the main or v3 branches. I created a local workaround using data_folder/output/success.json to manually update the blacklist between runs. Suggestion: make this a built-in feature with configuration options.

• "How did you hear about us": When this is a text field sometimes has odd values. Suggestion: default to "LinkedIn" or provide a config option for default.

• Bot-Like Answers: Some answers suggest bot-generated responses, e.g., "I don't have information for First Last..." (where "First Last" is the applicant). Suggestion: add sanity checks for response length, especially for short expected answers. Optionally log questionable responses or introduce a "strict" mode.

• Job History: It would be helpful to have a detailed log of all job applications and responses for reference, especially when contacted for an interview. This would also help with debugging.

Bug:

• Cover Letter Bug: Textbox cover letters are reused, sometimes including the wrong company name. Suggestion: prevent reuse of form fields, especially for cover letters.

• stringWidth Bug In linkedin_easy_applier.py, utils.stringWidth is referenced but doesn't exist. I fixed it locally by importing stringWidth from reportlab.pdfbase.pdfmetrics.

• Hiring Manager Questions: Inadequate handling of these, though I lack specific examples at the moment to share.