prior-auth-packet-builder
Build a concise prior authorization packet from local case files and payer policy docs.
소스 정보
- 저장소
- openai/openai-agents-python
- 최근 소스 활동
- 2026년 7월 31일 22:58
- 감지된 SKILL.md 언어
- 영어
- 스타
- 29,842
- 포크
- 4,845
설치 방법
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
소스 파일 검토
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
SKILL.md 표시 중
SKILL.md
소스 지침 · 읽기 전용 미리보기- name
- prior-auth-packet-builder
- description
- Build a concise prior authorization packet from local case files and payer policy docs.
# Prior Auth Packet Builder
Use this skill when a case requires prior authorization review, referral validation, imaging review, or payer-specific policy checks.
## Workflow
1. Inspect `case/scenario.json` and `case/transcript.txt`.
2. Search `policies/` for payer, prior auth, referral, imaging, and PPO guidance:
- Run the preferred search and fallback as one shell command:
`rg -n -i 'prior authorization|prior-auth|imaging|referral|billing|PPO|Blue Cross' policies ||
grep -RniE 'prior authorization|prior-auth|imaging|referral|billing|PPO|Blue Cross' policies`.
- An `rg` launcher or bootstrap failure is not evidence that there are no policy matches.
3. Read only the most relevant policy files.
4. Create `output/policy_findings.md` with these exact headings:
- `## Case summary`
- `## Matched policy files`
- `## Prior authorization`
- `## Referral`
- `## Missing information`
Cite each matched policy by its filename.
5. Create `output/human_review_checklist.md` with:
- what a human reviewer should verify
- what to tell the patient
- what queue should own the case
6. Call `finalize_policy_packet` only after both artifacts exist.
## Rules
- Use targeted searches over broad file reads.
- Use the targeted `grep -RniE` fallback only when `rg` cannot complete the search.
- Only cite policy files you actually inspected.
- Keep outputs concise and operational.
- The workflow is complete only when `finalize_policy_packet` succeeds.
- If referral status is pending and prior auth is unclear, recommend human review.
GitHub에서 보기