| name | resume-panel |
| description | Fan out a 12-persona review panel against will_cygan_resume.typ in parallel, with work-experience/*.md supplied as the source-of-truth evidence base, and produce a consolidated dashboard with consensus verdict, dissent highlights, and deduplicated top issues. Use when the user wants a comprehensive multi-angle resume review, asks for a "panel review", "full review", "multi-angle review", or wants every reviewer to weigh in at once. Covers ATS parsing, recruiter triage, hiring manager, technical screener, bar raiser, staff IC peer, leveling committee, skip-level exec, domain SME, typography, future-self skeptic, and career coach perspectives. |
| allowed-tools | Read, Grep, Glob, Agent |
Resume Panel Review
Run a parallel fan-out of all 12 reviewer sub-agents defined in .claude/agents/ against will_cygan_resume.typ, with work-experience/*.md supplied as the evidence base, then synthesize their outputs into a single dashboard.
Source-of-truth hierarchy
Every reviewer must treat these three inputs with distinct roles. Brief them explicitly in the spawn prompt — do not assume the persona definition covers it.
will_cygan_resume.typ — the artifact under review. This is what recruiters, ATSs, and hiring managers actually see. All verdicts and line references land here.work-experience/01-linkedin-swe.md and work-experience/02-linkedin-sr-swe.md — the source of truth for every LinkedIn role. Contains full Context / Role / Actions / Impact / Tech / Status logs and raw evidence for every claim. work-experience/99-personal-projects.md covers the Projects section. work-experience/ACRONYMS.md glosses LinkedIn-internal terms.advice/*.md — methodology (STAR/XYZ, quantification, readability). Agents already know to cite these; you do not need to re-brief.
Reviewers must use (2) in three directions:
- Inflation / under-support check. Any resume claim whose scope, number, or scale is not corroborated by
work-experience/ is a flag. Example: a QPS, revenue, or row-count figure on the resume should match the work-experience entry; unverifiable numbers are a bar-raiser / future-self-skeptic issue.
- Buried-lead check. Strong, well-quantified material in
work-experience/ (especially high-RS projects) that is missing from, weakly represented on, or out-ranked by weaker material on will_cygan_resume.typ is a flag. This is usually a hiring-manager, skip-level-exec, staff-ic-peer, or career-coach issue.
- Relevance Score calibration. Each project in
work-experience/*.md carries a **Relevance:** N/5 — … field documented in work-experience/RELEVANCE.md. Use the score as a prior: RS 5/4 projects missing from the resume are high-priority buried leads; RS 2/1 projects on the resume are wasted real estate. If a reviewer disagrees with a score (inflated or under-rated), it must say so — the score is human judgment being calibrated against, not gospel.
Process
-
Fan out. In a single message, launch all 12 sub-agents in parallel using the Agent tool — one call per persona. Use the spawn prompt in REFERENCE.md § Spawn Prompt exactly as written; it encodes the source-of-truth hierarchy above.
-
Collect. As each returns, record its verdict, confidence, top issue, and line references in the summary table. Line refs land in will_cygan_resume.typ; evidence/counter-evidence citations land in work-experience/*.md.
-
Tally consensus. Count the verdicts across all 12. Identify the dominant verdict and any dissenters.
-
Highlight dissent. For each persona whose verdict disagrees with the majority, quote their strongest reason. Dissent is high-signal — surface it, don't average it away. If all 12 agree, say so explicitly.
-
Deduplicate issues. Merge top issues from all 12 into a ranked list of 10 max. Weight by cross-persona support (more personas raising the same issue = higher priority) and severity. A single strong technical-depth concern from domain-sme-systems — or a single well-evidenced buried-lead from hiring-manager citing a specific work-experience/ bullet — may still rank above a broader but shallower complaint.
-
Emit the dashboard. Use the output format in REFERENCE.md § Dashboard Format. The Consolidated Top Issues section must cite work-experience/ line refs whenever an issue is framed as inflation or a buried lead.
Notes
- The 12 sub-agents are defined in
.claude/agents/ and already know which /advice/*.md files to cite. You do not need to brief them on methodology — but you DO need to brief them on the source-of-truth hierarchy above, since most personas' own definitions do not reference work-experience/.
- Expect sub-agent outputs to be noisy. Synthesize aggressively — 10 consolidated issues is the cap, not the target.
- Recommend a follow-up skill or command at the end of the dashboard based on the consolidated issues (e.g.,
/resume-panel-focus, /resume-debate, or a tactical skill from resume-optimizer).
See REFERENCE.md for the full persona roster, the exact spawn prompt, and the output template.
