메뉴
Talk to Paul Jialiang Wu's portfolio agent over A2A — grounded, honest Q&A, claim verification, and role-fit, machine-to-machine.
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/wjlgatech/agentic-portfolio-public --skill agentic-portfolio-a2a명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
SKILL.md 표시 중
SOC 직업 분류 기준
| name | agentic-portfolio-a2a |
| description | Talk to Paul Jialiang Wu's portfolio agent over A2A — grounded, honest Q&A, claim verification, and role-fit, machine-to-machine. |
| homepage | https://agentic-portfolio-lovat.vercel.app |
| protocol | a2a |
This portfolio is agent-native: another AI agent (a recruiter's screener, a sourcing bot, a collaborator's agent) can discover it and talk to it directly, machine-to-machine, over Google's A2A (Agent2Agent) protocol. Authored in the spirit of CLI-Anything — making all software agent-native.
Every answer is grounded and honest: it comes only from the portfolio's evidence + the verified "Receipts" claims. Private projects share a high-level highlight only. Unprovable claims (employment, degrees) come back "unverified — needs an external source", never faked.
curl https://agentic-portfolio-lovat.vercel.app/.well-known/agent-card.json
# (legacy clients: /.well-known/agent.json — both work)
The card advertises the JSON-RPC endpoint (url), capabilities (sync; streaming:false),
and the skills below.
POST to the card's url (/api/a2a). Use message/send (legacy tasks/send also works).
The response is a completed A2A Task whose status.message / artifacts carry the answer.
curl -s -X POST https://agentic-portfolio-lovat.vercel.app/api/a2a \
-H 'Content-Type: application/json' \
-d '{
"jsonrpc": "2.0",
"id": "1",
"method": "message/send",
"params": {
"message": {
"role": "user",
"parts": [{ "kind": "text", "text": "Does he have production experience building agent-verification tooling?" }],
"metadata": { "skill": "verify_claim" }
}
}
}'
skill (pass as params.message.metadata.skill) | Use |
|---|---|
ask_candidate (default) | Any grounded question about his work, projects, skills, mission, values. |
verify_claim | Check a claim → VERDICT: corroborated|partial|unverified|contradicted + citations. |
role_fit | Give a role/JD → matched strengths (with evidence) + honest gaps + FIT: rating. |
No auth — public read access by design. Be a good citizen: one question per call, keep text short.