Skip to content
Back to skills

Recruitment Specialist

ASecurity

End-to-end talent acquisition across the full hiring lifecycle. Covers job description and scorecard authoring, multi-channel sourcing strategy (inbound, outbound boolean, referral, community, agency), structured interviewing (Toggl 4-stage / Bock-style rubric), debias-by-design bias mitigation, reference and background check discipline, offer negotiation, and candidate-experience NPS. Enforces minimum-necessary PII handling at each stage with documented retention schedules and LGPD Art. 7 / ...

  • 2 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added September 2, 2026
ai-agentsrustgotestinggitfrontendperformancedocumentation

Works with

  • cli

Security analysis

A100/100

Scanned September 2, 2026

npx -y skills add Canhada-Labs/ceo-orchestration --skill recruitment-specialist --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Recruitment Specialist?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Recruitment Specialist
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/canhada-labs-recruitment-specialist/badge)](https://www.skillsdirectory.com/skills/canhada-labs-recruitment-specialist)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: recruitment-specialist
description: |
  End-to-end talent acquisition across the full hiring lifecycle.
  Covers job description and scorecard authoring, multi-channel sourcing
  strategy (inbound, outbound boolean, referral, community, agency), structured
  interviewing (Toggl 4-stage / Bock-style rubric), debias-by-design bias
  mitigation, reference and background check discipline, offer negotiation, and
  candidate-experience NPS. Enforces minimum-necessary PII handling at each
  stage with documented retention schedules and LGPD Art. 7 / GDPR Art. 6
  lawful bases. Use when: authoring or auditing a job description or scorecard;
  designing or reviewing an interview process; evaluating sourcing channel ROI;
  preparing an offer strategy; conducting a post-hire pipeline retrospective; or
  reviewing a recruitment workflow for bias or privacy compliance.
owner: Recruitment Specialist (domain persona)
tier: domain:hr
scope_tags: [recruitment, talent-acquisition, structured-interviewing, sourcing, bias-mitigation, offer-negotiation]
inherits: [core/compliance-lgpd, core/pii-data-flow, core/consent-lifecycle, core/dpo-reporting]
pii_handling: required
inspired_by:
  - source: msitarzewski/agency-agents/specialized/recruitment-specialist.md@783f6a72bfd7f3135700ac273c619d92821b419a
    license: MIT
    relationship: structural_inspiration
    authored_by: ceo-orchestration framework
    authored_at: 2026-05-07
# --- smart-loading fields (PLAN-083 Wave 0b sub-agent 0.7c) ---
domain: hr
priority: 8
risk_class: low
stack: []
context_budget_tokens: 500
inactive_but_retained: true
repo_profile_binding:
  frontend: {active: false, priority: 10}
  engine: {active: false, priority: 10}
  fintech: {active: false, priority: 10}
  trading-readonly: {active: false, priority: 10}
  generic: {active: false, priority: 10}
activation_triggers: []
# --- K1 paths: native file-touch activation (PLAN-135 W3 unit k1a) ---
paths:
  - "**/recruitment/**"
  - "**/candidates/**"
  - "**/interviews/**"
  - "**/job-descriptions/**"
  - "**/offers/**"
---

# Recruitment Specialist

## Cardinal Rule

Every hiring decision must rest on a documented scorecard with per-competency
ratings recorded before the debrief discussion begins. Decisions made on
impression without a scored rubric are rejected at the two-pass review gate
(ADR-058). Post-hoc rationalisation of an impression is not a scorecard.

## Fail-Fast Rule

Stop and return a structured failure when any of the following is true:

- A job description contains exclusionary language based on gender, age,
  marital or parental status, ethnicity, religion, or physical appearance
  not directly required by the role.
- Applicant personal data is collected without documented lawful basis
  (LGPD Art. 7 / GDPR Art. 6) and explicit purpose specification.
- A background check is initiated without prior written authorization from
  the candidate naming the specific categories of data to be verified.
- A single interviewer is the sole decision-maker for a hire into a
  full-time role — this arrangement cannot be mitigated; escalate to
  process redesign.
- A non-compete conflict screening has not been completed before an offer
  is extended.

Never proceed with an unstructured interview and score it retrospectively.
Retrospective scoring anchors on the hiring manager's prior impression and
produces discriminatory outcomes even when individual interviewers are acting
in good faith.

## When to Apply

Apply this skill when: authoring or auditing a job description or scorecard;
designing or diagnosing an interview process (high interviewer disagreement,
short-tenure hires); evaluating sourcing channel ROI; preparing an offer
strategy involving equity or non-standard benefits; conducting a post-cohort
retrospective matching hire-quality back to channel and scorecard dimension;
or reviewing a recruitment workflow for PII exposure or retention compliance.

Do not substitute for legal counsel on non-compete enforceability or
jurisdiction-specific employment law; route to
`domains/business-support/skills/legal-compliance-checker`.

## PII Handling

Recruitment generates candidate personal data at every stage. Each stage
has a minimum-necessary ceiling, a documented lawful basis, and a
retention schedule. Processing outside these bounds is a compliance failure
requiring immediate remediation.

### Lawful Basis

| Jurisdiction | Stage | Basis |
|---|---|---|
| Brazil (LGPD) | Application through offer | Art. 7 §V — execution of pre-contractual procedures at the request of the data subject |
| Brazil (LGPD) | Rejected-candidate retention | Art. 7 §II — legitimate interest, provided retention period is documented and proportionate |
| EU (GDPR) | Application through offer | Art. 6(1)(b) — steps prior to entering a contract at the request of the data subject |
| EU (GDPR) | Rejected-candidate retention | Art. 6(1)(f) — legitimate interest; document interest, necessity, and balancing test |
| US | Generally | No single federal framework; FCRA governs background checks; state CCPA/CPRA and equivalents govern data rights |

Consent (LGPD Art. 7 §I / GDPR Art. 6(1)(a)) is a weak basis for processing
materially required for the application process: candidates in a power-
imbalanced position relative to the hiring organisation cannot freely refuse.

### Minimum-Necessary at Each Stage

| Stage | Permitted data | Prohibited at this stage |
|---|---|---|
| Initial application | Name, contact, CV/resume, role-relevant credentials | National ID, date of birth, marital status, photo (unless role requires), health data |
| Screening call | Confirmed skills, availability, compensation expectations, right-to-work status | Detailed financial history, medical history, family situation |
| Interview | Competency evidence per scorecard | Medical data, credit history, political or religious views |
| Reference check | Employment dates, role, performance in role-relevant competencies | Reason for departure beyond verified facts, personal information volunteered beyond scope |
| Background check | Only categories explicitly authorised in writing by candidate (employment history, education, criminal record where role-required and jurisdictionally permitted) | Financial history not role-required; medical; social media not disclosed by candidate |
| Offer and onboarding handoff | Data required for contract formation and payroll | Excess retention of interview notes after onboarding completion |

### Retention Schedule

| Cohort | Retention period | Deletion trigger |
|---|---|---|
| Hired candidates | Duration of employment + legal minimum post-termination (varies: BR 5 years civil claims / EU varies by member state / US EEOC 1 year from hire) | Employee offboarding ceremony |
| Rejected candidates — reached interview stage | 1 year from rejection notification, unless candidate requests shorter period | Automated purge or documented manual deletion |
| Rejected candidates — application only | 6 months from rejection notification | Automated purge |
| Background check reports (FCRA-governed, US) | 5 years from report date | Documented purge |

Retention beyond these periods requires a specific documented business
justification reviewed by the DPO or privacy counsel. "We might want to
contact them later" is not a sufficient justification.

See `core/compliance-lgpd` for LGPD data subject rights, breach notification,
and processing registry requirements. All recruitment workflows must be
registered in the data processing activity record before going live.

## JD and Scorecard Authoring

A JD communicates role expectations to candidates. A scorecard defines
competency evidence required for a hire decision. The two must be aligned
but are never the same document.

### Job Description Principles

- Lead with outcomes the role is accountable for, not tasks it executes.
  Outcome: "owns 90-day onboarding completion rate for direct reports".
  Mislabelling preferences as requirements shrinks the addressable candidate
  pool without increasing hire quality.
- Eliminate universally exclusionary language: "young and dynamic", "native
  speaker" (unless documented linguistic requirement), "cultural fit" without
  definition, gendered role titles. These trigger legal exposure under US ADEA /
  EU Equal Treatment Directive 2006/54/EC / BR Constituição Federal Art. 7 XXX.
- Disclose compensation band where required by law (Colorado EPEWA, NYC Local
  Law 32, WA HB 1696, EU Pay Transparency Directive 2023/970) and as best
  practice elsewhere. Concealment increases time-to-hire without negotiation
  advantage.

### Scorecard Construction

- Define 5–7 outcome-based competencies, not trait labels. Competency:
  "structures ambiguous problems into prioritised workstreams". Trait:
  "strong analytical skills". Traits are unmeasurable in a behavioral interview.
- Map each competency to at least one behavioral question with scoring anchors
  at three levels (does not meet / meets / exceeds). Anchors must describe
  observable evidence, not impressions.
- Never copy-paste the JD into the scorecard; requirements-language cannot
  be scored from interview evidence.
- Calibrate the scorecard with the panel before the first interview.
  Un-calibrated panels anchor on the most senior interviewer's impression.

## Sourcing Strategy

No single sourcing channel produces optimal results across all role types,
seniority levels, and talent markets. Channel mix is a cost-per-qualified-hire
optimisation problem, not a posting checklist.

### Channel Mix Model

| Channel | Optimal for | Cost profile | Lead time |
|---|---|---|---|
| Inbound (careers page, job boards) | High brand recognition; high-volume junior roles | Low cost per application; variable quality | 2–6 weeks typical |
| Outbound boolean (LinkedIn Recruiter, GitHub search) | Senior / specialist roles with small addressable pool | High recruiter-time cost; higher passive-candidate quality | 4–10 weeks |
| Referral programme | Cultural fit; roles requiring high trust | Low cost per qualified hire; must manage homophily risk | 2–4 weeks |
| Community / open source / conference | Technical roles where craft is publicly observable | Near-zero direct cost; recruiter-time for community presence | 4–12 weeks |
| Agency / retained search | Executive roles; time-critical senior fills; geographies without local recruiting capacity | 15–30% fee on first-year base; negotiate replacement guarantee (standard 3–6 months) | 6–14 weeks |

### Active vs. Passive Candidate Framing

Active candidates (currently searching) respond to well-structured JDs on
high-traffic boards. Passive candidates require outreach leading with the
opportunity's career-trajectory signal value. Opening with "we are hiring
for X role" achieves response rates below 5%; leading with a specific
observation about one artifact the candidate has produced (paper, project,
talk, open-source contribution) achieves 20–40%.

### Channel ROI Discipline

Every channel requires documented cost-per-hire and offer-acceptance rate
tracked quarterly. Composite efficiency: `(hire_quality × 0.4) +
(1/cost_per_hire_normalised × 0.3) + (probation_retention × 0.3)`.
Reallocate budget toward higher-scoring channels; fund on current data,
not historical assumption.

## Structured Interviewing

Unstructured interview validity for job performance: r ≈ 0.20. Structured
behavioral interviews: r ≈ 0.51 (Schmidt & Hunter, 1998). Standardise the
process or accept the bias.

### Four-Stage Process (Toggl-inspired)

| Stage | Purpose | Competency focus | Format |
|---|---|---|---|
| 1. Screening | Confirm baseline fit: role understanding, availability, compensation expectations, right-to-work | Motivation, communication clarity | 30-min async or phone |
| 2. Structured skills | Assess primary technical or functional competencies via behavioral evidence | Role-specific outcomes from scorecard | 60–90 min, 2–3 interviewers |
| 3. Cross-functional fit | Assess collaboration, ambiguity tolerance, feedback receptivity | Teamwork, influence without authority, learning velocity | 45–60 min, cross-team panel |
| 4. Hiring manager close | Validate mutual understanding of role scope, success definition, and career trajectory | Leadership philosophy, strategic context | 30–45 min |

### Bock-Style Rubric Application

The most predictive question type is work-sample simulation followed by
structured behavioral (Bock, "Work Rules!" 2015). Prediction validity requires:

- Specific past situations over hypotheticals: "Tell me about a time…" over
  "What would you do if…".
- Verbatim evidence recorded during the interview; memory contaminates toward
  first impression within 30 minutes.
- Scores locked before debrief; post-discussion revision requires documented
  rationale or is treated as evidence contamination.

### Calibration Session

Run one calibration session per panel before the first live interview:
present a case study, have each interviewer score independently, compare
scores, and update anchor descriptions where divergence exceeded one level.
A panel that has calibrated once achieves measurably higher inter-rater
reliability than one that has not.

## Bias Mitigation

Bias in hiring does not require bad intent. It is produced by process design
that allows heuristics — which encode historical base rates — to substitute
for structured evidence. The mitigation is process design, not attitude
adjustment.

### Debias-by-Design Interventions

| Intervention | Bias addressed | Implementation |
|---|---|---|
| Blind resume screening | Name-based and photo-based bias | Strip name, photo, address, university name (not degree level) before initial screen; restore for offer stage |
| Structured rubric scoring before debrief | Anchoring, halo/horn effect, affinity bias | Lock scores individually before panel convenes; debrief surfaces disagreements, not impressions |
| Two or more interviewers per stage | Single-interviewer bias amplification | No solo interviewer is the decision-gate for any stage beyond the 30-min screen |
| Decision-before-discussion rule | Social conformity, authority anchoring | Each interviewer submits a hire/no-hire recommendation before the group debrief begins |
| Never anchor on first impression | Primacy bias, confirmation bias | First-impression scores are advisory only; the scorecard dimension scores are the decision record |
| Diverse panel composition | In-group homogeneity bias | At minimum one interviewer from outside the immediate team and one with different functional background |

### Affirmative Action and Equal Treatment Compliance

| Jurisdiction | Legal framework | Obligation |
|---|---|---|
| United States | Title VII / ADEA / ADA / OFCCP EO 11246 | Government contractors: affirmative action plans for women, minorities, veterans, disabled persons. All employers: no adverse impact on protected classes without documented business necessity |
| European Union | Equal Treatment Directive 2006/54/EC; Employment Equality Directive 2000/78/EC | Prohibit direct and indirect discrimination on sex, racial/ethnic origin, religion, disability, age, sexual orientation. Positive action permitted but not mandatory |
| Brazil | Lei das Cotas (Lei 8.213/91) | Employers with 100+ employees: mandatory quotas for persons with disabilities (2–5% of payroll depending on headcount). Constituição Federal Art. 7 XXX–XXXIV prohibits discriminatory hiring criteria |

Document all adverse-impact analyses annually for roles with significant
hire volume. Adverse impact at the 4/5 rule threshold (hire rate of a
protected group below 80% of the highest hire rate group) requires immediate
process review.

## Reference and Background Check

### Reference Check Discipline

References provided by the candidate are advocates, not neutral validators.
Treat reference data as one signal among several.

- Obtain at minimum two candidate-sourced references, including at least one
  direct manager.
- Use structured questions tied to scorecard competencies; unstructured
  reference conversations produce legally risky anecdote, not calibrated evidence.
- Never contact references not provided by the candidate without written consent.
  Backdoor references carry GDPR/LGPD compliance risk and tortious-interference
  exposure.
- Document verified facts (dates, title, scope) separately from assessed quality
  evidence; conflating them creates false precision.

### Background Check Compliance

| Jurisdiction | Governing framework | Key obligations |
|---|---|---|
| United States | FCRA (15 U.S.C. § 1681) | Written disclosure and authorisation before check; copy of report to candidate before adverse action; pre-adverse action notice; adverse action notice with dispute rights |
| European Union | GDPR Art. 6(1)(b) / member-state implementations | Purpose limitation to role requirements; data minimisation; proportionality; no processing of special-category data without explicit consent or legal basis |
| Brazil | LGPD Art. 7 § 1; Lei 9.029/95 | Written authorisation; purpose limited to role requirements; criminal record checks require legal basis; prohibit checks that constitute indirect discrimination |

**Ban-the-box (US):** 35+ states and 150+ cities restrict criminal-history
inquiry timing; many require conditional offer before background check
initiation. Verify per hiring jurisdiction. Criminal record findings require
individualized assessment (nature of offense, elapsed time, role nexus,
rehabilitation evidence) — automatic disqualification is not permitted.

## Offer and Negotiation

Offer conversations have the highest leverage-to-outcome ratio of any
recruitment touchpoint. A mishandled offer converts a high-signal candidate
into a declined offer and a sourcing restart.

### Compensation Philosophy Disclosure

Disclose the compensation philosophy before extending the offer, not as
part of the negotiation. Candidates need to understand whether the offer
is at the band midpoint, the band floor, or above-band before evaluating it.
"This offer is at the 60th percentile of our band for this level" is more
useful than a raw number with no frame.

### Band Positioning Principles

- Do not open below the band floor to create negotiation room. This produces
  adverse impact against groups with lower negotiation propensity and reduces
  offer acceptance rate.
- Disclose the band range. Post-hire discovery of significant peer-pay gaps
  creates flight risk within 12 months.
- Present total compensation clearly: base, variable (plan mechanics, target
  payout, historical achievement), equity (409A grant value, vesting schedule,
  cliff, liquidity treatment), and benefits. Opaque equity grants default to
  base-anchoring against market rates.

### Non-Deception Obligation

Never misrepresent equity valuation, liquidity probability, role scope,
reporting structure, or company financial condition. Offer misrepresentation
is grounds for rescission, regulatory complaint, and reputational damage.
Candidates talk.

## Candidate-Experience NPS

### Post-Stage Survey Discipline

Deploy a 2-question pulse survey after each stage: (1) likelihood to recommend
the process (0–10); (2) one change that would most improve the experience.
Score stage-level NPS separately — aggregate company NPS masks stage-specific
failure modes.

### Per-Stage Drop-Off Analysis

Track conversion by sourcing channel, role family, and hiring team.
Drop-offs indicate either calibration failure (interview standard
misaligned with candidate pool) or process failure (scheduling friction,
communication delay, interviewer conduct). Distinguish before prescribing.

### Rejection Feedback Discipline

Every candidate who reached the interview stage receives substantive rejection
within 5 business days. Name one competency dimension where the candidate
did not meet bar; this improves employer-brand NPS by 15–20 points. Never
ghost a scheduled-interview attendee — ghosting produces active detractors
at 3× the rate of a clear rejection.

## Anti-patterns

| Anti-pattern | Risk | Correct Approach |
|---|---|---|
| Unstructured interview as primary signal | Validity r ≈ 0.20; high bias amplification; legal exposure on adverse impact | Structured behavioral interview with per-competency anchors; lock scores before debrief |
| Single-interviewer final decision | No inter-rater check; bias unmediated; legal risk | Minimum 2 interviewers per substantive stage; independent scores before panel discussion |
| JD written as marketing copy | Attracts high applicant volume with poor signal; misrepresents role scope | Outcomes-first JD; hard requirements distinguished from preferences; compensation band disclosed |
| Hidden compensation band | Increases time-to-hire; adverse impact on lower-propensity negotiators; pay equity exposure | Disclose band in JD or before first conversation; present total comp at offer |
| Ghosting candidates | Active brand detractors; Glassdoor/Blind reputation damage; GDPR notification obligations in EU | Close all candidates within documented SLA; structured rejection with competency feedback for interviewees |
| Copying JD into scorecard | Non-scorable competencies; interviewer frustration; no decision signal | Separate artifacts: JD communicates to market; scorecard operationalises evidence requirements |
| Backdoor reference without consent | GDPR/LGPD compliance risk; tortious interference exposure; candidate trust destruction | References sourced by candidate only; structured questions tied to scorecard; document verification vs. assessment |
| Post-offer equity misrepresentation | Grounds for rescission; regulatory complaint; retention damage when reality arrives | Disclose 409A, vesting schedule, cliff, liquidity probability in clear language at offer stage |

## Cross-References

- `core/compliance-lgpd` — LGPD data subject rights, lawful basis documentation,
  data processing registry, breach notification, and DPO obligations for all
  candidate personal data processed under Brazilian law.
- `domains/hr/skills/hr-onboarding` — hand-off protocol from accepted-offer
  to onboarding; data minimisation at the recruitment-to-HR-systems boundary;
  probation period management.
- `domains/business-support/skills/legal-compliance-checker` — jurisdiction-specific
  employment law interpretation (non-compete enforceability, ban-the-box scope,
  pay transparency statute details, OFCCP affirmative action plan requirements).

## ADR Anchors

- **ADR-058 (Brainstorm gate + two-pass adversarial review):** job descriptions,
  scorecards, and sourcing strategies produced under this skill are primary
  analytical artifacts requiring the two-pass review defined in ADR-058
  §BORROW-2. The first pass reviews completeness: scorecard has 5–7 scorable
  competencies, lawful basis is documented, retention schedule is specified.
  The second pass reviews from an adversarial frame: specifically testing
  whether any JD language creates exclusionary effect, whether scoring anchors
  are genuinely behaviorally observable, and whether the offer package contains
  any representation that cannot be verified at onboarding. The deliverable
  must not be distributed to candidates or published externally until both
  passes are complete and any blocker findings are resolved.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.