Job-Hunter-Linkedin-Skill-Hermes-OpenClaw-ClaudeCode-Comet/skills/hr-candidate-hunter.md

# HR Candidate Hunter
**Agentic candidate sourcing, LinkedIn outreach, and talent pipeline management.**
Stop waiting for applicants. The best candidates aren't applying — they're being found. This skill turns HR into a talent intelligence operation.

## Philosophy
Traditional recruiting is post-and-pray. Modern recruiting is **identify, engage, convert**. The best hires are passive candidates who don't know they want to leave yet.

The model:
1. **Profile** — Define the ideal candidate beyond the job description
2. **Source** — Multi-platform search with Boolean mastery
3. **Engage** — Personalized outreach that gets replies (not ignores)
4. **Qualify** — Screen efficiently without killing the candidate experience
5. **Nurture** — Build a pipeline for roles that aren't open yet

## Input Required
- Job description or role requirements
- Company info (stage, culture, perks, location, remote policy)
- Compensation range
- Timeline and urgency
- Any existing candidate pool or CRM data

## Workflow

### Phase 1: Ideal Candidate Profile
```
Before searching, define WHO you're looking for:

Must-Have (deal-breakers if missing):
  - [Required skill/experience 1]
  - [Required skill/experience 2]
  - [Required skill/experience 3]

Strong Preferences (flexible for exceptional candidates):
  - [Preferred background]
  - [Preferred industry experience]
  - [Preferred location/working model]

Nice-to-Have (differentiators, not requirements):
  - [Bonus skill]
  - [Relevant certification]
  - [Specific company background]

Cultural Signals (harder to search, easier to screen):
  - Grit/perseverance indicators
  - Intellectual curiosity signals
  - Collaboration evidence
  - Growth mindset markers

Anti-patterns (red flags to filter OUT):
  - Job hopping without growth (4+ roles in 5 years with no title progression)
  - Skills listed without context
  - Misaligned compensation expectations
  - Values misalignment with company culture
```

### Phase 2: Agentic LinkedIn Sourcing
```
LinkedIn is the primary sourcing channel. Here's how to use it systematically:

Boolean Search Strings:

For Software Engineers:
"(("software engineer" OR "software developer" OR SDE)
("distributed systems" OR "microservices" OR "system design")
(Python OR Java OR Go OR Rust))
NOT ("recruiter" OR "HR" OR "human resources")
("series a" OR "series b" OR "startup" OR "scale-up")"

For Product Managers:
"(("product manager" OR "senior product manager")
("roadmap" OR "product strategy" OR "PRD")
("B2B" OR "SaaS" OR "platform"))
NOT ("associate product manager" OR "product marketing")"

For Designers:
"(("product designer" OR "UX designer" OR "senior designer")
("design systems" OR "Figma" OR "user research")
("mobile" OR "web" OR "enterprise"))"

Search Filters to Apply:
- Title: Exact or related titles (current, not past)
- Location: City, region, or "Open to remote"
- Company: Current company size or type
- School: Target universities (if relevant)
- Years of experience: Appropriate range
- "Open to work" filter: Passive + active candidates

Sourcing cadence:
- 50-100 profiles reviewed per role per day
- 20-30 connection requests per day (LinkedIn limit)
- 10-15 personalized messages per day
- Track all outreach in candidate tracker
```

### Phase 3: Multi-Platform Sourcing
```
Don't rely on LinkedIn alone. Source across platforms:

GitHub (for engineers):
- Search repos using target languages/frameworks
- Check contribution quality, not just count
- Look for maintainers of relevant open-source projects
- Reach out via email found in commit history or profile

Twitter/X (for builders and creators):
- Search industry hashtags (#buildinpublic, #engineering)
- Find people sharing technical content
- Engage with their content before reaching out
- DM for warm introduction

Stack Overflow (for engineers):
- Top answerers in relevant tags
- People asking thoughtful questions
- Community reputation as signal

Industry Communities:
- Discord/Slack communities in target domain
- Meetup speakers and organizers
- Conference speakers and workshop leaders
- Open source community members

Portfolio Sites:
- Dribbble/Behance (designers)
- Product Hunt makers (product people)
- Substack/Medium authors (thought leaders)
- YouTube creators (educators/builders)
```

### Phase 4: Outreach Engine
```
Outreach templates that get replies (NOT generic recruiter spam):

LinkedIn Connection Request (300 char limit):
"Hi [Name], your work on [specific project/post/skill] caught my
eye. We're building [what] at [Company] and looking for someone
with [specific expertise]. Would love to share details."

First Message (after connection):
"Thanks for connecting, [Name].

I reached out because [specific reason — NOT "I came across your
profile"]. We're hiring a [Role] at [Company] and your background
in [specific skill/experience] is exactly what we need.

Quick details:
- Role: [Title]
- Team: [Size, what they build]
- Tech: [Stack]
- Comp: [Range if shareable]
- Remote: [Policy]

Would you be open to a 15-min chat this week? No pressure either
way — happy to just share what we're building."

Cold Email (for GitHub/portfolio sourced candidates):
Subject: [Specific compliment — NOT "exciting opportunity"]

"Hi [Name],

I found your work on [specific repo/project/article] and was
impressed by [specific detail]. That's rare to see.

I'm [Name] at [Company]. We're [brief company context — 1 line].

We're looking for a [Role] to [specific challenge they'd solve].
Based on your [specific experience], I think you'd thrive here.

If you're at all curious, I'd love to share more. No commitment.

[Your Name]
[Link to company or role page]"

Follow-up cadence:
- Day 0: Initial outreach
- Day 3: Gentle follow-up (different angle)
- Day 7: Value-add follow-up (share relevant content)
- Day 14: Final check-in (respect the silence)

Reply rate benchmarks:
- Generic recruiter message: 5-10%
- Personalized message: 25-40%
- Warm referral: 50-70%
- Engaged-with-their-content-first: 35-50%
```

### Phase 5: Screening & Qualification
```
Before the hiring manager sees anyone, pre-qualify:

15-Minute Phone Screen Script:
1. "Tell me about what you're working on now." (2 min)
   → Are they articulate? Passionate? Honest about challenges?

2. "What are you looking for in your next role?" (2 min)
   → Align motivations with what you offer

3. "Walk me through a recent project you're proud of." (5 min)
   → Depth of knowledge, ownership, impact language

4. "What's your ideal timeline for a move?" (1 min)
   → Are they actively looking, passively open, or just curious?

5. "What compensation range would make you excited?" (1 min)
   → Align expectations before investing more time

6. "Any questions about the role or company?" (4 min)
   → Their questions reveal their priorities

Green flags from screen:
- Asks thoughtful questions about the team/challenges
- Speaks in specifics, not generalities
- Talks about impact, not just activity
- Honest about gaps and growth areas

Red flags from screen:
- Vague about their own work
- Only motivated by money or title
- Speaks negatively about current/past employers
- No questions about the role
- Compensation expectations wildly misaligned
```

### Phase 6: Pipeline Building
```
Build a living talent pipeline, not a one-time search:

Pipeline structure:
  Hot (ready to hire now):    5-10 candidates
  Warm (open in 1-3 months):  15-20 candidates
  Cold (future potential):     30-50 candidates
  Nurture (long-term):         50+ candidates

Nurture tactics:
- Monthly company update email (achievements, culture, openings)
- Invite to company events, meetups, AMAs
- Share relevant content (blog posts, talks, launches)
- Congratulate on career milestones (promotions, launches)
- Re-engage quarterly with personalized check-ins

Pipeline metrics to track:
- Source quality (which channels produce best candidates)
- Response rate by outreach type
- Time-to-engage (first contact → first response)
- Conversion rate (outreach → screen → interview → offer)
- Pipeline coverage (candidates per open role)
```

## Agentic Search Automation
```
Automated sourcing workflows you can run with AI tools:

Daily sourcing loop:
1. Run Boolean searches across LinkedIn (50 profiles)
2. Score each profile against ideal candidate profile (AI-assisted)
3. Generate personalized outreach for top 15 matches
4. Send connection requests with custom notes
5. Log all activity in candidate tracker
6. Flag high-priority candidates for immediate outreach

Weekly pipeline review:
1. Check response rates and adjust messaging
2. Update pipeline status for all candidates
3. Identify drop-off points in the funnel
4. Source fresh candidates to maintain pipeline health
5. Report pipeline status to hiring manager

AI-assisted candidate scoring:
For each candidate, rate (1-5):
- Skill match: How well do they match must-haves?
- Experience level: Appropriate seniority?
- Culture signal: Evidence of cultural alignment?
- Growth trajectory: Upward trend in career?
- Availability signal: Any indicators of openness to move?
- Compensation alignment: Likely within range?
- Quality of work: Portfolio/GitHub/writing quality?

Score 25+: Fast-track outreach
Score 20-24: Standard outreach
Score 15-19: Nurture pipeline
Score <15: Pass
```

## Integration with Other Skills
- **hr-job-description-forge**: Job descriptions inform candidate profiles
- **hr-interview-designer**: Screening feeds into structured interviews
- **hr-offer-architect**: Comp research informs sourcing strategy
- **hr-talent-pipeline**: Pipeline management and long-term talent strategy
- **hr-culture-architect**: Cultural values inform candidate filtering

## Files
- `memory/hr/candidate-tracker-[role].md` — Active candidate tracking per role
- `memory/hr/sourcing-plan-[role].md` — Sourcing strategy per role
- `memory/hr/outreach-templates.md` — Tested outreach templates and response rates
- `memory/hr/pipeline-[department].md` — Department talent pipeline
- `memory/hr/sourcing-metrics.md` — Channel performance and conversion data