Application Q&A

Questions ATS auto-fill couldn't answer · type once, future identicals auto-fill.

ℹ How this page works (click to expand)

What's on the page

  • Pending — questions an active application is waiting on. Type an answer + hit Save (or press Enter) → the paused application can resume.
  • Recently answered — green chips that pop up when an answer just got saved (auto-clear after 30 min).
  • Already answered (N) — every Q&A entry the agent currently has. Filter, edit, or delete inline.

Inline edit / delete

  • Click Edit → question + answer become editable. Save writes to data/applicant_profile.yaml via an atomic file-locked write. Cancel discards.
  • Click Delete → the entry is removed from the YAML. Future occurrences of that question land in pending + email you.

Manual YAML edits also work

  • Open data/applicant_profile.yaml in any editor. Save your change.
  • The next time the agent calls lookup(), it sees the file's new mtime → invalidates the cache → re-reads. No restart needed.

Token answers (e.g. $${experience.notice_period_days})

  • Some answers reference profile fields via ${dotted.path} tokens. Edit experience.notice_period_days at the top of the YAML and every "notice / join / availability" answer updates everywhere automatically.
  • ${fx.expected_in_USD}-style tokens get converted live via frankfurter.app from the base INR values (24-hour cache). Don't replace these tokens with literal numbers unless you specifically want to break the link.
  • The grey "→ resolves to: ..." line under each answer shows what the agent will actually type.

Email reply path

  • Unanswered questions also send you a [CV-Agent] email asking for the answer.
  • Reply to the email with just the answer text. The watcher (python -m src.apply.qna_email_watcher) parses the reply, updates the profile, and a green chip pops up on this page within ~5 minutes.

Four ways to add an answer (all feed the same YAML)

  1. Type it in the Pending table here
  2. Click Edit on a row in the Already answered table
  3. Reply to the [CV-Agent] email when one arrives
  4. Hand-edit data/applicant_profile.yaml directly

Every answer you save here gets appended to data/applicant_profile.yaml. The next time the agent sees a question that matches, it auto-fills. You can also edit the YAML directly — file changes are auto-detected (mtime check), no restart needed. Or reply directly to the [CV-Agent] emails — the watcher (python -m src.apply.qna_email_watcher) parses the reply and updates the profile automatically.