A skill for B2B sales workflows: lead qualification, safe product Q&A, demo booking, CRM note writing, and human handoff.

• Runs natural lead conversations without jumping into hard selling.
• Collects company size, industry, pain points, budget, decision maker, and timeline.
• Answers product questions only from approved knowledge-base evidence.
• Requires email, timezone, and demo purpose before booking demos.
• Writes CRM notes with pain points, stage, next action, and confidence level.
• Escalates high-risk legal, security, contract, and custom quote requests.

b2b-sales-agent/
  SKILL.md
  agents/openai.yaml
  references/
  scripts/
scripts/
tests/
examples/
docs/superpowers/

Copy the b2b-sales-agent directory into your Codex skills directory:

Copy-Item -Recurse .\b2b-sales-agent "$env:USERPROFILE\.codex\skills\b2b-sales-agent"

Then invoke it with:

Use $b2b-sales-agent to qualify this lead and decide the next sales action.

python -m unittest discover -s tests -v
python scripts/validate_skill.py .\b2b-sales-agent
python scripts/validate_test_cases.py tests\fixtures\improved_agent_test_cases.json
python scripts/evaluate_dialogue_policy.py tests\fixtures\problem_dialogues.json

The skill expects the host agent environment to expose these tools:

• get_lead_context(lead_id)
• search_knowledge_base(query)
• check_calendar(timezone, duration_minutes)
• book_demo(lead_id, slot_id, attendee_email, summary)
• write_crm_note(lead_id, summary, qualification_level, next_action)
• handoff_to_human(lead_id, reason, urgency)

The local scripts do not call live systems. They provide deterministic helpers and tests for the skill's workflow rules.

lchang1125@163.com