career-engine-intake

Intake Pipeline

Registry: this pipeline is listed in the Pipeline Registry in skills/career-engine/SKILL.md. Actions owned by another pipeline's registry row are out of scope here — route to that pipeline instead of improvising.

This skill has two modes with different procedures — Steps 0 through 0.9 below are Notion-fetch mode's procedure only. Inline mode is a separate, much shorter flow that ends after the bullet below; it never reaches Step 0, never creates $PIPE (Step 0.4 explicitly skips it), and never touches Steps 0.6 through 0.9 (priority queue, the batch coach spawn, the Coach Output Check, or Notion writeback) — there is no Notion row to write back to and no batch to queue.

• Inline mode — the user provides a URL or JD text directly in chat. No Notion fetch, no $PIPE, no queue.md. Run JD acquisition (the Step 0.5 fetch ladder, applied to the single provided URL/JD) on that one role, then spawn career-coach with Option 1 — Inline (not Option 2 — Option 2 is exclusively for the batch/Notion path, per its own "Always invoked... for Needs Research roles" scope statement), passing the fetched content directly in the spawn prompt along with CAREER_DATA=${CAREER_DATA} (resolved at Step −0.5 or by the orchestrator preflight, same as every other coach spawn in this file) — the coach needs it to read 01/02/03 for this role's analysis exactly as it does in the batch path; omitting it here is not a smaller ask, it is a missing required input. Deliver the coach's response conversationally — no Notion writeback, no further steps. Use when the user says something like "coach me on this role" and pastes a URL or JD, outside of the batch Notion-fetch flow.
• Notion-fetch mode — queries the Notion database for Needs Research roles, runs JD acquisition and the career coach (Option 2) for each, writes all results to Notion, and updates Status to Researched. This is the standard "run intake" path, and the one Steps 0 through 0.9 below describe.

The career coach runs for every role that is not already coach-complete. If all roles in the queue are coach-complete, the coach spawn is skipped and the pipeline proceeds directly to Step 0.9 using existing values. The goal of this pipeline is to give the career coach complete information — full JD data for every role — before it makes any prioritization or writing decisions.

career-data data root (R-37). The personal-data files — 01-writing-rules.md, 02-professional-background.md, 03-framework.md, linkedin-profile.md, pipeline-preferences.json, delivered-letters/, and the user's .dotx — load from ${CAREER_DATA}/references/, the path the orchestrator resolves in its career-data discovery preflight. Every other file (self-checks, REFERENCES.md, skill docs, default .dotx templates) stays on ${CLAUDE_PLUGIN_ROOT}. If ${CAREER_DATA} is not set (direct or standalone invocation outside the orchestrator), locate the career-data skill yourself, confirm career-data-marker.json, and apply the orchestrator's healthy / damaged / absent outcomes before reading. A configured user's missing career-data is a hard stop — never silently fall back to blank templates.