| name | board |
| description | Application tracker board. Read ~/.dear-hiring-manager/applications.md and render a status board (grouped by blocked/skipped/filled/submitted/rejected/interview/offer with counts and a response-rate summary), and update an application's status + date when the user reports progress (e.g. "mark Transcend as interview"). Use for /board or whenever the user wants to see or update their job-application pipeline. |
Tracker board
Read and maintain the application pipeline in ~/.dear-hiring-manager/applications.md (the flat-file
tracker). Two jobs: show the board and update a status.
Show
- Read
applications.md. If it's missing or has no rows, say so and point to
/dear-hiring-manager:apply or :batch.
- Render a board grouped by
status, in pipeline order:
blocked → skipped → filled → submitted → rejected → interview → offer.
- A count per column header.
- One line per application under its status:
company — role (fit N · applied <date> · N flags).
- A one-line summary: totals, and response rate =
(interview + offer) / (submitted + rejected + interview + offer) — i.e. responses over everything
ever submitted (status is single-valued, so an advanced app has left the submitted bucket). Show
n/a when that denominator is 0.
- Surface what's actionable:
blocked apps (CAPTCHA / unreadable form) → need your manual action; list them first.filled apps not yet submitted → parked, awaiting the human's Submit.submitted apps with no status change in >14 days → consider a follow-up.
Update
When the user reports progress (e.g. "Transcend → interview", "Airbnb rejected"):
- Find the matching row in
applications.md by company/role.
- Set its
status and append a dated note in the flags/notes column (e.g. interview 2026-07-20).
Never delete history — append.
- Re-render the board.
Rules
- Flat-file only;
applications.md is the single source of truth, human-readable Markdown.
- Status lifecycle: in-progress → blocked | skipped | filled → submitted → rejected | interview | offer.
- Never fabricate a status — only reflect what the human reports or what
apply/batch recorded.
