AgentStack
SKILL verified MIT Self-run

Decision Maker Finder

skill-xuxchloris-hermes-export-skills-decision-maker-finder · by Xuxchloris

Use when a trade agent needs decision-maker clues, buyer roles, purchasing manager signals, sourcing contacts, category manager hints, contact-page evidence, or email validation status

No reviews yet
0 installs
8 views
0.0% view→install

Install

$ agentstack add skill-xuxchloris-hermes-export-skills-decision-maker-finder

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Decision Maker Finder? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Decision Maker Finder

Overview

Find role-level contact clues from company pages and optional enrichment API results. The core rule is to output decision-maker evidence with source_url, role, and email_status while also forcing a separate email and phone search across checked pages.

When to Use

Use this skill when the user asks who to contact, whether a company has purchasing clues, or how to enrich a prospect with decision-maker signals.

Use company-research first when the company fit is still unknown.

Inputs

  • Company website
  • Company name and country, if available
  • Decision-maker roles from MARKET.yaml
  • Contact enrichment settings from DISCOVERY.yaml
  • Optional output from prospect-list-enrichment

Outputs

{
  "website": "",
  "pages_checked": [],
  "contact_search": {
    "email_result": "found|没有",
    "phone_result": "found|没有",
    "emails": [],
    "phones": []
  },
  "candidates": [
    {
      "name": "",
      "role": "",
      "email": "",
      "email_status": "missing|invalid_format|format_valid|domain_match|api_verified",
      "phone": "",
      "phone_status": "missing|found",
      "confidence": "low|medium|high",
      "source_url": "",
      "evidence": ""
    }
  ],
  "review_notes": []
}

Procedure

  1. Run python tools/decision_maker_finder.py --website --output when file output is needed.
  2. If the input is only a company name or company link, resolve the official website when possible before contact analysis.
  3. Check company pages such as homepage, about, contact, team, catalog, and product pages.
  4. Always run official website contact search across every checked page for email and phone values, even when no role clue is found.
  5. If email or phone is not found, output email_result: "没有" or phone_result: "没有" in contact_search.
  6. Look for role clues such as Owner, Founder, Purchasing Manager, Sourcing Manager, Category Manager, Procurement Manager, and Buyer.
  7. Validate email syntax and mark whether the email domain appears company-level.
  8. If a phone appears near a role clue, include it with phone_status.
  9. If DISCOVERY.yaml has a configured contact enrichment API, use its schema and record API results as additional evidence.
  10. Return role, name if visible, email if visible, phone if visible, confidence, source URL, and evidence text.
  11. Pass useful candidates into prospect-scoring as decision-maker access evidence.

Verification

  • Every candidate includes a role and source URL.
  • contact_search is present even when there are no candidates.
  • Official website contact search runs when a company name, website, or company link is available.
  • email_status is present for every candidate.
  • Missing email or phone search results are written as 没有.
  • A role clue is not described as confirmed purchase intent.
  • API keys are read from environment variables, not written into output files.
  • Low-confidence clues remain review items.

Common Mistakes

| Mistake | Fix | | --- | --- | | Treating a role mention as a named contact | Separate role, name, and evidence | | Calling a generic email verified | Use format_valid unless stronger evidence exists | | Leaving phone or email blank after search | Put 没有 in contact_search result fields | | Dropping source pages | Keep source_url for every candidate | | Sending outreach automatically | Return candidates for human review |

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.