application-qa

name	application-qa
description	Generates copy-pasteable answers for job application portal questions. Takes a role reference (#N or company name), fetches the JD, and produces answers grounded in your stories, identity, and skills matrix — each sourced to specific evidence. Say "answer questions for #68" or "answer Harvey Director" or "portal questions for Kadence". Also handles custom questions when you paste them.
triggers	["aq","answer questions for","answer questions","application questions","portal questions","app questions for"]

Application QA — Career OS Skill

Purpose

Application portals ask questions beyond the resume. "Why this company?", "Describe a time you...", "What's your management style?" — these require tailored answers grounded in YOUR specific stories and the specific JD.

This skill generates those answers so you can copy-paste them into any application portal. Each answer is sourced to specific evidence files so you can verify and edit before submitting.

Output Format

Always start with:

━━━ Career OS: Application QA — {Company} {Role} ━━━

How to Invoke

• answer questions for #68 — generate standard answers for role #68
• answer Harvey Director — fuzzy resolve, then generate
• answer questions for Kadence — same flow
• answer #68: "Why do you want to work here?" — answer a specific question
• Paste custom questions and say answer these for #68 — answer whatever the portal asks

---

DATA ARCHITECTURE

Inputs

Source	Path	What It Provides
Role context	Resolved via pipeline-query.py --lookup	Company, role, score, JD URL, resume track
JD content	Fetched from JD URL (WebFetch) or cached in career-intelligence/projects/job-search/scans/	Requirements, responsibilities, company values
Identity	brain/identity/identity.md	Career narrative, values, motivation, "why"
Stories	brain/stories/*.md	Evidence for behavioral/situational answers
Skills matrix	brain/identity/skills-matrix.md	Technical proficiency evidence
Match scoring	career-intelligence/projects/job-search/job-pipeline-match-tracker.json	Category scores, gap analysis, match rationale
People	network/people/*.md	Warm contacts (for "how did you hear about us")
JD Alignment Framework	brain/projects/jd-alignment-framework.md	Track definitions and match evidence — used for grounding portal answers in alignment data
Resume Generation Guide	brain/projects/resume-generation-guide.md	Generation context — ensures portal answers are coherent with resume narrative

Outputs

Output	What It Contains
Console output	Formatted Q&A with copy-pasteable answers and source citations

This skill does NOT write to files. It generates answers for the user to copy-paste into application portals.

---

BEHAVIOR: Generate Standard Answers

Step 1: Resolve Role

Accept #N or fuzzy name. Resolve using:

python3 $CAREER_HOME/~/.career-os-state/scripts/pipeline-query.py --lookup "{input}" --format json

If ambiguous, present disambiguation (A/B/C scheme).

If the role context was already provided by apply-dashboard (pre-resolved), skip the lookup.

Step 2: Load Context

Read these files (lazy-load — only what's needed):

1. JD content — Try in order: a. Fetch from JD URL via WebFetch (if URL is available and valid) b. Read from scan cache: career-intelligence/projects/job-search/scans/{batch-date}/ matching company/role c. Read from match tracker detailed scoring section for this role d. If none available, work from the role title + company research only

2. Identity — brain/identity/identity.md (always read)

3. Stories index — brain/stories/STORY_INDEX.md (read index, then selectively read 3-5 most relevant stories based on JD requirements)

4. Match scoring — Read the detailed scoring section for this role from the match tracker (provides category scores and gap analysis)

5. People — Check network/people/ for contacts at this company (for "how did you hear about us" and referral mentions)

Step 3: Generate Standard Q&A

Generate answers for these common portal questions. Skip any that don't apply to this specific role/company:

Q1: "Why do you want to work at [Company]?"

• Sources: identity.md (career motivation) + company mission/values from JD + match rationale from scoring
• Approach: Lead with genuine alignment between your career arc and their mission. Reference specific company initiatives or products.
• Length: 150-200 words

Q2: "Describe a relevant accomplishment"

• Sources: Best-fit story from stories/ matched to the JD's primary emphasis
• Approach: Pick the story with highest relevance to the JD's top-weighted category. Use STAR format with concrete metrics.
• Story selection priority: Match story theme to JD emphasis (AI, scale, leadership, 0→1)
• Length: 200-300 words

Q3: "What's your management/leadership style?"

• Sources: identity.md + leadership stories
• Approach: Describe coaching philosophy and leadership approach. Ground in specific examples.
• Length: 150-200 words

Q4: "Why are you leaving your current role?"

• Sources: identity.md (career transition narrative)
• Approach: Frame positively — seeking the next challenge. Keep professional and forward-looking.
• Length: 100-150 words

Q5: "How did you hear about this role?"

• Sources: People files, scan reports, warm contacts
• Approach: If there's a warm contact → mention the referral. If not → mention your targeted search.
• Length: 1-2 sentences

Q6: "What's your experience with [specific technology from JD]?"

• Sources: skills-matrix.md + relevant stories
• Approach: Map the technology to your proficiency level. If proficient, cite projects. If learnable, frame adjacent experience.
• Length: 100-200 words (only generate if JD calls out specific tech)

Step 4: Format Output

━━━ Career OS: Application QA — {Company} {Role} ━━━

Q1: Why do you want to work at {Company}?
━━━
[Answer text — 150-200 words]

Sources: identity.md, stories/{slug}.md, match-tracker #{N}

---

Q2: Describe a relevant accomplishment
━━━
[Answer text — STAR format, 200-300 words]

Sources: stories/{slug}.md

---

[...continue for each applicable question...]

━━━
→ Copy-paste these into the application portal
→ Edit to add personal touches before submitting
→ Say "applied to #{N}" when done to update your pipeline

---

BEHAVIOR: Custom Questions

When the user pastes specific portal questions:

1. Resolve the role reference
2. Load context (same as standard flow)
3. For each custom question:
   • Identify the best-fit story/evidence
   • Generate a tailored answer
   • Cite sources
4. Output in the same format as standard Q&A

---

GUARDRAILS

• Never fabricate experience. Every claim must trace to a story file, skills-matrix entry, or identity statement. If there's no evidence for something, frame it as a learning opportunity, not existing expertise.

• Match the company's tone. If the JD is formal, keep answers professional. If the JD is casual, match that energy while staying substantive.

• Respect the gap analysis. If the match tracker shows a gap in a category, don't oversell that area. Acknowledge adjacent experience and frame the learning path.

• Length discipline. Portal text fields often have character limits. Keep answers within the specified word counts. Concise and compelling beats long and generic.

---

Scope Boundary (WO-044)

This skill answers portal application questions (text fields). It does NOT generate cover letters — that is cover-letter's job. If a portal question field asks "paste your cover letter" or "why are you interested," redirect:

"This field wants a cover letter. Say 'cover letter for {Company}' to generate one as DOCX + PDF."

Do not generate the cover letter inline in an answer field. Cover letters are a separate deliverable with their own QA gates, file format, and opt-in gate (see cover-letter/SKILL.md).

INTERACTION WITH OTHER SKILLS

Skill	Relationship
apply-dashboard	Upstream — routes "answer questions for #N" here with pre-resolved context
job-match-scorer	Upstream — scoring rationale informs answer framing
cover-letter	Peer — owns cover-letter generation and QA gates
resume-engine	Peer — answers complement the resume (don't contradict resume framing)
apply-tracker	Downstream — "applied to #N" after submitting answers
story-capture	Upstream — stories are the evidence base for all answers