job-trackerlisted

Use this skill whenever the user wants to log, update, or review job applications — e.g. "add this application to my tracker," "log that I applied to Acme for the Staff Engineer role," "update the status on my Beta Corp application to Interview," "what applications need a follow-up," "show me my job search pipeline," "I heard back from X, mark it as rejected," or any request to record or check on job-search progress over time. This skill maintains one persistent tracker file across sessions — on first use it creates the tracker; on later use it reads the existing tracker before adding, updating, or querying it, and never overwrites or drops existing entries. Trigger this even if the user gives only partial details (e.g. just a company name and status change) — ask for the minimum needed to identify or create a row rather than guessing. No file attachment is required to use this skill; it manages its own persistent output file.
nikhilvdev/job-search-pack · ★ 0 · AI & Automation · score 73

Install: claude install-skill nikhilvdev/job-search-pack

# Job Tracker Maintains a single running tracker of job applications across sessions — add an application, update its status, or query the pipeline, without ever losing data from a previous session. ## Why this skill is different from the rest of the pack Every other skill in this pack produces one document from one conversation. This one is stateful: it reads an existing file, changes a small part of it, and writes the whole thing back — over and over, across many separate conversations. The core discipline here isn't "produce good content," it's "never silently lose or corrupt what's already there." Always read the full existing tracker before writing anything, and always write back the full table, not just the part you touched. ## Required inputs Inputs depend on which operation is being requested — there's no fixed upfront list: - **Add an application**: company + role, at minimum. Status defaults to "Applied" unless the user says otherwise. Date applied, posting URL, contact, and notes are optional — include them if given, don't block on them if not. - **Update an application**: enough to identify the existing row — company + role, or company alone if there's no ambiguity — plus what's changing. - **Query the tracker**: no new inputs needed beyond the user's intent (e.g. "what's overdue for a follow-up," "show me everything at Offer stage"). If a request is ambiguous about which row it refers to, ask rather than guess — see Step 4. ## Step 1: Locate or create the