Greenhouse Bundle

Turn Claude Code into a Greenhouse-native hiring analyst. Search Harvest pipelines, score applicants against a role brief (with scorecards where your kit allows), turn interview notes into structured feedback, export multi-job workbooks, draft jobs from intake notes, and generate leadership-ready reports—so hiring managers spend less time clicking and more time deciding.

Why this bundle exists

Greenhouse is the system of record for many high-growth teams, but bulk sense-making (compare candidates across roles, apply one rubric everywhere, refresh exports for finance or DEI reporting) still happens in spreadsheets. This bundle keeps applications and scorecards in Harvest while Claude handles interpretation, ranking, and narrative—auditable and repeatable from your project.

How you benefit

• Consistent evaluation — One pasted brief can drive score-candidates across a stage or job so interviewers see the same dimensions.
• Faster stakeholder packs — export-pipeline and generate-report produce Excel and narratives executives can read without Harvest seats for every viewer.
• Less copy-paste from intake calls — create-requisition drafts structured jobs from notes; you refine in Greenhouse as usual.
• Feedback that survives — write-feedback structures raw notes into scorecard-aligned feedback where your process supports it.

Example prompts

When Claude has access to Greenhouse through this bundle’s API integration (with your keys and project configured), you can ask in plain language—for example:

• Find applications stuck in Internal Review for job ID 456 and summarise why they might be blocked.
• Score everyone in the onsite loop for this role against our scorecard template and prepare Harvest-ready feedback.
• Export every open requisition’s pipeline to separate sheets for the quarterly DEI report.
• Generate a two-page narrative on pipeline health and scorecard usage for leadership this month.

Full power (easy to miss)

• Ask for a chain in one instruction: e.g. Screening for role X → score against this brief → export full pipeline → weekly metrics narrative.
• Pair with sourcing bundles (Apollo, Hunter, Lusha, ContactOut) for net-new talent; use this bundle when candidates already exist in Harvest and you need velocity and quality on review.

What you get

Digital download: a ZIP you unzip into your project. Configure GREENHOUSE_API_KEY, GREENHOUSE_USER_ID (for On-Behalf-Of on writes), and ANTHROPIC_API_KEY.

Included skills

Skill	What it does
search-candidates	Pipeline search by job/stage or multi-wave talent search
score-candidates	AI-score applicants and submit Harvest scorecards
write-feedback	Raw notes → structured scorecard on an application
export-pipeline	Export pipeline to Excel (one sheet per job)
create-requisition	Create a job with department and office
generate-report	Metrics plus Claude narrative (applications + scorecards)

How it works

1. Unzip the bundle into your Claude Code project (see buyer setup in the archive).
2. Add credentials to a shared .env at the project root (see Requirements).
3. Ask Claude in plain English — e.g. “Search Screening for senior engineers on this job, score them against this brief, then export the pipeline.”

Requirements

• GREENHOUSE_API_KEY — Harvest API key (Basic auth username, blank password)
• GREENHOUSE_USER_ID — Greenhouse user id for On-Behalf-Of on writes and activity-feed notes
• ANTHROPIC_API_KEY — Claude API access

Compatibility

Works with macOS and Windows (run.sh / run.ps1). Designed for Claude Code and similar assistants.

Support

Technical reference: CLAUDE.md and per-skill README files in the bundle. Release history at the bottom of this page mirrors CHANGELOG.md so you can read version notes before purchase.

For administrators (API permissions)

Public Harvest v1 documentation emphasizes GET scorecard patterns. This bundle submits scorecards via POST /scorecards with a JSON body aligned to your interview kit (application_id, interview_step_id, overall_recommendation, attributes, questions). Grant the matching permission on your API key and validate payloads against your organization’s kits if you receive HTTP 422 responses.

Release history

Below is the same content as CHANGELOG.md in this bundle, so you can see version history before purchase. After download, both files ship together in the ZIP.

Greenhouse Bundle Changelog

1.0.2 (2026-04-30)

Add .env.example at the bundle root documenting environment variables for this integration (copy to .env). No skill behavior changes.

1.0.1 (2026-04-30)

Maintenance release: refreshed buyer-facing PRODUCT.md release history (mirrored from CHANGELOG.md) and rebuilt the distribution ZIP. No skill behavior changes.

1.0.0 (2026-04-29)

Initial release.

Skills included

• search-candidates: Pipeline search and talent search with AI ranking (Harvest API)
• score-candidates: AI scoring against job brief, submits Harvest scorecards
• write-feedback: Structured interview feedback as Greenhouse scorecards
• export-pipeline: Full pipeline export to multi-tab Excel
• create-requisition: Job creation from intake notes (POST /jobs)
• generate-report: Pipeline analytics with scorecards and Claude narrative insights