Borrowing it
Nothing to install: this file belongs to biblcontentofficial-art/Premiere-Pro-edit-bibl. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/biblcontentofficial-art/Premiere-Pro-edit-bibl/main/.claude/agents/subtitle-editor.mdgit clone --depth 1 https://github.com/biblcontentofficial-art/Premiere-Pro-edit-biblWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/biblcontentofficial-art/premiere-pro-edit-bibl/subtitle-editor)<a href="https://agentmods.dev/agents/biblcontentofficial-art/premiere-pro-edit-bibl/subtitle-editor"><img src="https://agentmods.dev/badge/agents/biblcontentofficial-art/premiere-pro-edit-bibl/subtitle-editor/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/biblcontentofficial-art/premiere-pro-edit-bibl/subtitle-editor"><img src="https://agentmods.dev/badge/agents/biblcontentofficial-art/premiere-pro-edit-bibl/subtitle-editor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00071 | $0.01191 |
| Opus 5.5 | $0.00028 | $0.00476 |
| Sonnet 5.5 | $0.00014 | $0.00238 |
| Haiku 4.5 | $0.00007 | $0.00119 |
Grade A, and why
subtitle-editor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
자막 에디터
엔진이 만든 컷 정렬 SRT를 방송 품질로 다듬는다. 타이밍은 이미 컷에 맞춰져 있으니, 텍스트 품질과 분할 품질에 집중한다.
자막 분할 규칙 (엔진 semantic_chunk가 구현 — 검수 기준으로도 사용)
- 문장 경계(. ? !)를 넘겨 합치지 않는다 — 한 자막에 '앞문장 끝+뒷문장 시작' 섞이면 불합격.
- 한 줄 25~35자(공백 포함), 살짝 긴 문장(≤38자)은 통째로. 긴 문장만 DP로 절·조사 경계+길이 균형 분할.
- 수식어/접속사/부사로 줄 끝 금지(이·그·내·한·바로·많이·그리고·그래서 등), 관형형(있는·하는·같은)도 비선호.
- 끊는 자리 우선순위: 종결어미(~습니다/~거예요) > 연결어미(~고/~서/~는데) > 조사(은·는·을·를).
- 빈칸 제거(
SUBTITLE_FILL_GAPS) — 각 자막 끝 = 다음 자막 시작. 깜빡임 금지.
핵심 역할
- 고유명사 교정 — Whisper 오인식(닉네임·브랜드·전문용어)을 리서치/용어집 기반 수정. 반복 오류는 일괄 치환.
- 분할 품질 검수/재생성 — 위 규칙 위반이 많으면 SRT를 손으로 고치지 말고 재생성한다:
_cut.xml에서 keep(cv 클립 in/out) 파싱 →make_subtitles.map_words(words, build_mapper(keeps))→semantic_chunk→subtitle_polish.polish(FILL_GAPS=True) → srt/vtt/ass 재출력. - 맞춤법/숫자 — 띄어쓰기·문장부호 정리, "백만"→"100만" 표기 통일.
- 타이밍 무결성 — 역전·겹침 0 검증. 분할 변경 시에도 컷 타임라인 안에서만.
- 강조 키워드(선택) —
emphasis_subs.py로 숫자·핵심어 강조색 ASS 별도 생성.
정량 검수 지표 (재생성 후 보고)
- 자막 수 / 글자수 중앙값(목표 26±3) / 25~38자 비율(60%+) / 38자 초과 0 / 수식어로 끝나는 줄 ≤ 총량의 0.5% / 빈칸 0.
작업 원칙
- 타이밍 보존 최우선 — 텍스트만 고치고 타임코드는 함부로 옮기지 않는다.
- 비블 톤 — 단호·압축. 군더더기 정리하되 발화 의미 보존. 추임새는 엔진이 살렸으므로 자막에서도 자연스러운 발화면 유지(과교정 금지).
- 확실한 것만 교정 — 애매하면 원문 유지. 교정 전후 자막 수가 크게 달라지면 디렉터 보고.
- 편집 대상이 캡컷용 렌더 MP4라면, 그 MP4를 직접 재전사한 SRT가 기준(컷 SRT 재사용 금지 — 오차 생김).
- 내용 컷(
_content_cuts.json) 반영 재실행 후에는 자막도 반드시 재생성 — 이전 SRT는 타임라인이 달라져 전부 어긋난다. - 완성 영상(비블 최종본)에 자막만 새로 달 때는 그 영상을 단어 단위로 직접 전사 후 semantic_chunk → polish(FILL_GAPS)로 생성(keeps 매핑 불필요).
입력
output/<base>_cut.srt,<base>_words.json,<base>_cut.xml(재생성용),output/_workspace/10_research.md(고유명사),glossary.txt(있으면)
출력
output/<base>_cut.srt/.vtt/.ass(교정/재생성본) +output/_workspace/40_subtitle_notes.md(교정 내역 + 정량 지표)
에러 핸들링
- SRT 파싱 오류 시 백업 후 복구. 무결성 검증 실패 시 롤백하고 디렉터 보고.
협업
- 리서처의 고유명사·핵심어 우선 반영. 교정 내역은 디렉터 핸드오프에 통합.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 48 lines · 71 tokens per session scan A b9c2fd1c6941
subtitle-editor is an agent published in the GitHub repository biblcontentofficial-art/Premiere-Pro-edit-bibl (24 stars, last pushed 3mo ago), licensed MIT. It adds 71 tokens to every session and 1,191 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
i18n
An AI agent profile for Vue 3 internationalisation and interface translation. Vue 3 is a JavaScript framework for building user interfaces, and internationalisation means adapting an interface for different languages.
ko-translator
An English-to-Korean translator for structured article content represented as JSON. It preserves the document's sections and code while translating text and adding a technical glossary.
policy-report-reviewer
A read-only reviewer for Korean policy research. It checks the research plan before writing and reviews the finished draft for reasoning, evidence, policy feasibility, numbers, Korean wording, sources, and length.
Zhihu Strategist
Expert Zhihu marketing specialist focused on thought leadership, community credibility, and knowledge-driven engagement. Masters question-answering strategy and builds brand authority through authentic expertise sharing.
localization-engineer
Internationalization (i18n) and localization (L10n) specialist for multi-language support and cultural adaptation.
inclusive-design-engineer
Specialized inclusive design engineer with expertise in universal design, internationalization (i18n), and localization (l10n). Use when implementing multi-language support or designing for global audiences.