candidate-signal-analysis

Extract evidence-backed candidate facts and action cards from recruiter conversation screenshots and transcripts.

1|Updated Aug 3, 2026
One-click install
npx skills add https://github.com/getyak/talent-signal --skill candidate-signal-analysis-getyak
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: candidate-signal-analysis
Source: https://github.com/getyak/talent-signal/tree/main/.agents/skills/candidate-signal-analysis
Command: npx skills add https://github.com/getyak/talent-signal --skill candidate-signal-analysis-getyak

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Recruiters lose critical candidate context between conversations, forcing them to reconstruct details like deadlines, competing offers, and commitments from memory. This Skill turns conversation screenshots or transcripts into reviewable, evidence-cited facts and action cards so nothing decision-relevant slips through. ## Core Features & Use Cases - Evidence-Cited Fact Extraction: Pulls only explicit, decision-relevant facts (identity, availability, deadlines, preferences, commitments) with short source excerpts attached to each claim. - Reviewable Action Cards: Proposes create_contact, update_contact, or create_meeting cards that require human confirmation before any data changes. - Momentum Verdicts: Emits one verdict (advance, resolve_blocker, at_risk, or wait) plus a concrete next step with reason and timeframe. - Use Case: A recruiter pastes a chat screenshot where a candidate mentions a Wednesday decision deadline and a competing offer; the Skill returns an update_contact card citing the exact quote and an at_risk verdict with a next step to confirm remote-work policy within one business day. ## Quick Start Use candidate-signal-analysis to extract action cards and a next-step insight from this candidate conversation transcript.

Frequently Asked Questions about candidate-signal-analysis

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract candidate facts from a recruiting conversation transcript?▼

Provide the conversation transcript or screenshot plus optional recruiter context. The Skill extracts only explicit, decision-relevant facts such as deadlines, availability, and commitments, each with a short source excerpt and confidence level.

What action cards can be proposed from candidate conversations?▼

Only three card types are proposed: create_contact, update_contact, and create_meeting. Every card cites evidence and requires human confirmation before any data is created or changed.

How does candidate identity matching work with screenshots?▼

A contact is matched only when identity evidence is sufficient; otherwise a create_contact card is proposed. Speaker attribution is verified against visible labels, since left/right position alone does not establish a role.

Can the Skill predict whether a candidate will accept an offer?▼

No. It never claims a candidate will accept, decline, or churn. Risk is expressed only as an inference with supporting evidence, using verdicts like at_risk or wait.

What output format does the candidate analysis return?▼

It returns JSON with facts, action_cards, and an insight object containing a verdict, rationale, next step, and timeframe, as defined in references/output-contract.md. Ambiguous fields are marked low confidence rather than guessed.