Dossier Builder skill

Assembles complete research dossiers from all analysis components.

Script

dossier_builder.py

Compiles outputs from all other Jay skills into a standard research dossier. This is the final output sent to Rick for application positioning.

Call build_dossier(consultant_profile, job_posting, jd_analysis, staleness_result, tech_stack_validation, vendor_info, confidence_score) to generate complete dossier.

Dossier structure includes:

• •Dossier ID and timestamp
• •Recommendation (PASS/REVIEW/SKIP) with confidence score
• •Consultant profile (skills, experience, rate, visa)
• •Job details (title, company, posting age)
• •Vendor assessment (tier, ghost rate, track record)
• •Posting quality assessment (staleness, red flags)
• •Technology analysis (match percentage, stack validation)
• •Requirements analysis (MUST_HAVE, NICE_TO_HAVE, INFERRED)
• •Confidence breakdown (all scoring factors)
• •Risk assessment (high/moderate/low/critical)
• •Action items for Rick (resume optimization, cover letter focus, required validations)
• •Full metadata for auditability

Output goes to:

• •Rick: Full dossier for positioning and application building
• •Z: Submission request with risk level and recommendation

Rules

• •Never output a dossier with unresolved contradictions or tech stack validation failures.
• •Dossier is final research output before submission request to Z.
• •All dossiers are logged in application-history.jsonl for retrospective analysis.
• •Dossier builder runs last in Jay's workflow.