Back to library

Candidate source research

Researches candidate profiles against a role brief and produces an evidence-linked sourcing shortlist with fit gaps, provenance, and follow-up questions. Use for recruiting research, talent sourcing, executive search, candidate comparison, or pipeline qualification.

by adlass TemplatesVersion 1Uses adlass toolsUniversal

Published Aug 21, 2026 · Updated Aug 26, 2026

Helpful · 0View raw SKILL.md

Requirements

Add the process policy and field mapping for candidate_id, must_have, resume_evidence. Define thresholds, statuses, owners, and source precedence for employment_dates, source_provenance, follow_up.

Skill document

The full SKILL.md your agent reads and follows.

Candidate source research

Purpose

Turn role requirements and candidate source material into a comparable shortlist. Preserve what a profile actually says, distinguish verified employment evidence from self-reported claims, and surface the precise evidence still needed before a recruiter advances a person.

Scope

Assess role fit using title, employer, dates, scope, skills, certifications, location, work authorization, compensation expectation, availability, and stated achievements. Include candidate records, resumes, portfolios, referrals, and interview notes, plus the supplied hiring rubric.

Excluded: protected-characteristic inference, medical or personality assessment, candidate contact, background-check conclusions, ranking based on names or photos, and changes to an applicant-tracking system.

Data basis

  • Role brief with must-have, preferred, seniority, location, travel, and start-date requirements.
  • Candidate profile records, resumes, portfolios, referrals, and interview notes.
  • Hiring rubric, source provenance, and duplicate-candidate identifiers.
  • Optional run input for role, cohort, or shortlist size.

Result

A sourcing shortlist document and Candidate research notes sheet with one row per candidate and criterion, evidence status, citation, gap, and proposed follow-up question.

Quality criteria

  • Every scored criterion cites a resume section, profile field, portfolio item, or interview-note location.
  • Must-have criteria are separated from preferred criteria and missing evidence is not scored as failure.
  • Candidate names, current roles, dates, and source locations remain traceable.
  • Duplicate profiles are merged only when two identifiers support the match.
  • No protected trait is inferred or used in a comparison.

Instructions

Apply the supplied rubric exactly; if absent, use pass, evidence-needed, and not-matched without inventing weights. Verify chronology by checking overlapping employment dates and label simultaneous roles rather than resolving them. Treat a claim such as “led a global team” as unverified until team size, period, and source are present. Keep rejected candidates in the sheet with the unmet or unassessed criterion. Use neutral questions for gaps and never convert silence into a negative.

Adapt before use

  • Add the hiring rubric, role architecture, and approved evidence standards to the fixed scope.
  • Define privacy, retention, redaction, and access rules for candidate material.
  • Specify whether location, work authorization, compensation, or start date are hard filters.

Related skills

  • 360 feedback synthesis

    Synthesizes 360-degree survey responses into privacy-safe competency scores, rater-group themes, self-versus-other gaps, and a focused coaching agenda. Use for multi-rater reviews, leadership development, performance feedback, anonymized comment analysis, and competency calibration.

  • Absence pattern report

    Reports absence frequency, lost days, and recurring patterns from attendance records without inferring medical causes, and produces a cited HR report with a data-quality register. Use for leave analysis, attendance trends, capacity planning, absence reporting, or workforce reviews.

  • Candidate Data Extraction

    Extracts consistent candidate fields from CVs into a review-ready matrix, preserving source wording and marking absent or ambiguous information rather than inferring it. Use for resume parsing, applicant data normalization, recruitment data entry, or candidate profile preparation.

  • Candidate Interview Debrief

    Consolidates interview-loop notes into a candidate scorecard and decision memo, preserving question-level evidence, interviewer ratings, disagreement, and missing signals. Use after panel interviews, structured interview debriefs, hiring reviews, or candidate evaluation meetings.