market-brief

Scheduling (Claude Cowork)

Recommended: run weekly via Cowork /schedule (e.g. same day each week) so market signals stay fresh without daily noise. Example scheduled payload:

Run /career-navigator:market-brief (Career Navigator market-brief skill) for my configured {user_dir}.

Invoke the market-researcher agent to generate a current market intelligence brief grounded in the user's role targets and location preferences.

Important invocation/data-source rules:

• Use the exact agent name market-researcher (no aliases).
• Do not fail the skill because external web connectors are unavailable.
• Default to local evidence (profile.md, tracker.json, ExperienceLibrary.json, analyst norm tables, and AI report) unless the user explicitly asks for live web sourcing.
• If external sourcing is requested and unavailable, continue with local evidence and label confidence limits.

Workflow

1. Confirm required context

Read {user_dir}/CareerNavigator/profile.md.

If target roles are missing:

"I need your target role(s) to generate a market brief. Run /career-navigator:launch (or update CareerNavigator/profile.md) with your target roles first."

If location preferences are missing, continue but mark geographic confidence as limited.

Optionally read:

• {user_dir}/CareerNavigator/tracker.json for user-specific conversion/timeline signals
• {user_dir}/CareerNavigator/ExperienceLibrary.json for capability-fit context

2. Invoke market-researcher

Hand off to market-researcher with:

• The full CareerNavigator/profile.md
• The full CareerNavigator/tracker.json (if present)
• The full CareerNavigator/ExperienceLibrary.json (if present)
• Instruction to include all three required sections:
  1. Role demand trends
  2. AI/automation displacement outlook
  3. Geographic competitiveness signals

If the user asks about a specific role/geography, prioritize that in the brief and treat other profile targets as secondary.

If invocation fails due to agent naming/tooling:

• Retry once with the exact market-researcher name.
• If still failing, return a structured partial result from local files and clearly state that agent invocation failed.

3. Present the brief

Render the returned output in this structure:

**Market Brief** — {today's date}
Target role(s): {list}
Geography: {scope}
Confidence: {Preliminary / Directional / Moderate / High}

HIGHLIGHTS
- Demand: {1 line}
- AI outlook: {1 line}
- Geography: {1 line}

ROLE DEMAND TRENDS
{role-by-role demand posture and implications}

AI/AUTOMATION DISPLACEMENT OUTLOOK
{high-risk vs durable task clusters + positioning shift}

GEOGRAPHIC SIGNALS
{market-by-market competitiveness and timeline implications}

PRIORITIZED NEXT MOVES (30-60 days)
1. ...
2. ...
3. ...

Keep the brief practical: concise, evidence-based, and directly tied to search decisions.

4. Save the brief to disk

Write the rendered brief to {user_dir}/Market-Brief-{YYYY-MM-DD}.md in its own file-write call.

If the file tool fails: output the full brief in a fenced markdown block and tell the user to save manually; skip the index update if the file was not saved.

On success, update {user_dir}/CareerNavigator/artifacts-index.json in a separate call — append:

{
  "id": "{uuid}",
  "type": "market_brief",
  "filename": "Market-Brief-{YYYY-MM-DD}.md",
  "path": "{user_dir}/Market-Brief-{YYYY-MM-DD}.md",
  "target_company": null,
  "target_role": null,
  "date_created": "{today}",
  "source": "generated",
  "notes": "Market intelligence brief — {comma-separated target roles}"
}

5. Hand off to writer for PDF

Invoke the writer agent in market-brief-pdf mode, passing:

• The saved markdown file path from Step 4
• Today's date

The writer converts the markdown to PDF, saves it alongside the markdown in {user_dir}/, and presents the confirmed PDF path to the user as the completion deliverable for this task.

6. Suggest next step

Based on the top recommendation:

• If targeting shift is needed: suggest /career-navigator:search-jobs with refined role/location filters.
• If narrative shift is needed: suggest /career-navigator:assessment and then /career-navigator:tailor-resume.
• If execution cadence is the issue: suggest /career-navigator:follow-up and /career-navigator:track-application.