Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add bleakbelladonnals/asr-transcription-skills --skill asr-transcriptiongit clone --depth 1 https://github.com/bleakbelladonnals/asr-transcription-skillsWrote 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/skills/bleakbelladonnals/asr-transcription-skills/asr-transcription)<a href="https://agentmods.dev/skills/bleakbelladonnals/asr-transcription-skills/asr-transcription"><img src="https://agentmods.dev/badge/skills/bleakbelladonnals/asr-transcription-skills/asr-transcription/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/skills/bleakbelladonnals/asr-transcription-skills/asr-transcription"><img src="https://agentmods.dev/badge/skills/bleakbelladonnals/asr-transcription-skills/asr-transcription.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 | $0.00124 | $0.04393 |
| Opus 5 | $0.00062 | $0.02197 |
| Sonnet 5 | $0.00025 | $0.00879 |
| Haiku 4.5 | $0.00012 | $0.00439 |
Grade A, and why
asr-transcription 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 12d 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.
How it starts
The opening of the file, as written. The whole thing — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ASR 评测与转写管线
覆盖两类任务:
- 评测:对比不同录音硬件+转写引擎+纪要 AI 的质量(如飞书录音豆 vs 钉钉录音卡 vs 通义听悟),产出对比矩阵
- 转写:用在线 API(Gemini 免费层等)批量转写长音频(用户明确不装本地转写——在意磁盘空间)
核心方法论(评测类任务)
1. 双变量拆分(最重要)
贵硬件 = 采集端 + 转写端 + 总结端。必须分开测:
- 采集端(硬件):同期多设备录同一场内容,比录音文件本身
- 模型端(引擎):同一份高质量音频喂给不同转写服务,比转写文本
- 总结端(纪要):各家 AI 纪要单独比,并对照逐字稿回验
2. 控制变量
- 同一音频文件比引擎;同一场景同期录制比硬件
- 每段 5-15 分钟足够;总测试量控制在免费额度内
- 测试音频里埋"地雷":专业术语、易错英文名、数字——方便算保留率
3. 固定锚点测试集(跨引擎对比)
选 3-5 个有辨识度的句子/术语,每个引擎都查一遍,输出"通过/部分/失败"。本会话验证过的锚点类型:
- 英文产品名:
Claude Code(飞书妙记全程错成 Clang/Clash code,钉钉听记 8/8 对——同一声源铁证对比) - 专业词:
渐进式披露(各引擎差异极大) - 英文缩写:
MCP、anthropic、vibe coding - 幻觉检查:通义听悟免费版凭空编英文客服话术、无意义字母("AN""DK")、误听脏话
4. 纪要幻觉检测
AI 纪要会把"合理推测"写成事实(本会话抓到:回滚功能、定时触发/事件驱动、选题爬取都是逐字稿里没有的编造)。纪要必须对照逐字稿逐条回验,术语错误两家互不交叉。
5. 输出物
对比矩阵:音频质量表 + 转写质量表 + 纪要质量表 + 产品可用性(格式/导出/额度限制)+ 成本账。结论先行,数据可溯源。
评分维度量化(基于锚点计数折算,满分 100):专业术语准确率 / 整体信息保留 / 幻觉与乱码控制 / 中文口语流畅度 / 综合转写质量。
成本折算公式(用户口径:API 按量价 vs 会员折算价):
- 会员折算:
年费 ÷ 12 ÷ 月度额度分钟 × 60 ≈ 元/小时(钉钉 500+/年÷1300 分钟/月 ≈ 1.9 元/时;飞书 69 元/月÷~1300 分钟 ≈ 2.8-3.5 元/时) - API:新用户免费额度通常可覆盖个人月用量(本次 6 引擎实测总花费 0 元);超出按量以官网为准
- 报告注明"本次实测成本"与"长期折算成本"两列
HTML 报告模板:templates/asr_report_template.html——自包含 CSS 单文件(hero 头 + 九节:测试方法/音频质量/术语锚点表/综合评分/幻觉与信息丢失/成本/格式可用性/结论/踩坑记录),锚点表用「钉钉音频 / 飞书音频」双列(支撑"测转写必须先控制音频变量"论点)。用户要求的交付形态。
音频质量客观指标(ffmpeg,全片或抽段)
# 编码/时长/码率/声道
ffprobe -v quiet -show_format -show_streams in.mp3 | grep -E "duration|bit_rate|sample_rate|channels|codec_name"
# 响度/动态范围 (EBU R128)
ffmpeg -i in.mp3 -af ebur128 -f null - 2>&1 | grep -A8 Summary
# 峰值/RMS/削波(Peak 超 0dB = 削波风险)
ffmpeg -i in.mp3 -af astats -f null - 2>&1 | grep -E "RMS level|Peak level"
# 静音段(-38dB 阈值找不到静音 = AGC 抬底噪/压缩强)
ffmpeg -i in.mp3 -af silencedetect=noise=-38dB:d=2.5 -f null - 2>&1 | grep silence
# 双声道是否同源(RMS 几乎相同 = 非真立体声,对转写无额外信息)
ffmpeg -i in.ogg -af astats -f null - 2>&1 | grep -E "Channel|RMS level"
判读:响度 -27~-30 LUFS 正常;LRA 8 vs 14 = 压缩强弱;静音段存在 = 信号干净。
SRT 解析与统计(Python 标准库)
- 标准 SRT 正则:
r'(\d+)\n(\d{2}:\d{2}:\d{2},\d{3}) --> (\d{2}:\d{2}:\d{2},\d{3})\n(.*?)(?=\n\d+\n|\Z)'(re.S) - 统计:块数、总字数、均长/中位片段时长、<1s 占比(碎片化)、每分钟字数(中文 ~280-310 正常)
- 术语检查:对清洗后全文跑
re.findall(正确/错译 pattern)计数 - 清洗:钉钉导出文本行带
HH:MM:SS,mmm前缀(用户可选导出选项,不是 bug);通义听悟免费版导出是纯文本无时间戳(非标准 SRT)
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/asr-landscape-2026.md 4.5 KB
- references/asr-provider-quirks.md 6.5 KB
- references/bailian-hotwords-verified.md 5.2 KB
- references/feishu-vs-dingtalk-test-2026-08.md 3.0 KB
- scripts/bailian_sdk_transcribe.py 4.2 KB runs code
- scripts/gemini_transcribe.py 6.3 KB runs code
- templates/asr_report_template.html 4.9 KB
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.
- 12d ago First seen · 109 lines · 124 tokens per session scan A 80824e912394
asr-transcription is a skill published in the GitHub repository bleakbelladonnals/asr-transcription-skills (9 stars, last pushed 25d ago), licensed MIT. It adds 124 tokens to every session and 4,393 once invoked, about $0.0006 per session on Opus 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 2026-08-31.
Other skills, from other repositories
instrument-data-to-allotrope
Convert laboratory instrument output files (PDF, CSV, Excel, TXT) to Allotrope Simple Model (ASM) JSON format or flattened 2D CSV. Use this skill when scientists need to standardize instrument data for LIMS systems, data lakes, or downstream analysis. Supports auto-detection of instrument types. Outputs include full…
matlab
Build, review, migrate, and safely plan MATLAB or GNU Octave numerical workflows, including arrays, tabular/time data, tests, projects, graphics, MAT files, and explicit Python interoperability.
exploratory-data-analysis
Perform bounded, local exploratory analysis of explicitly supported scientific files. Use for redacted CSV/TSV/JSON profiles; optional NumPy, HDF5, FASTA/FASTQ, and basic image metadata inspection; missingness/leakage audits; outlier and transformation sensitivity; and rigorous EDA report scaffolds. Other domain…
phylogenetics
Build and analyze phylogenetic trees using MAFFT (multiple alignment), IQ-TREE 2 (maximum likelihood), and FastTree (fast NJ/ML). Visualize with ETE3 or FigTree. For evolutionary analysis, microbial genomics, viral phylodynamics, protein family analysis, and molecular clock studies.
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
mapping-to-snomed
Maps clinical concept spans extracted by OpenMed to SNOMED CT concepts through a USER-SUPPLIED terminology server (the user's own Ontoserver, Snowstorm, or UMLS/UTS), never a bundled vocabulary. Use when the user wants to code findings, disorders, procedures, body structures, or substances to SNOMED CT, run an ECL…