audio-tools-workflow

Use the RUDI audio-tools stack to transcribe, enrich, index, and query speech from local audio/video files, direct media URLs, supported video-page URLs, or base64 media input. Trigger when the user asks to process audio, video-linked speech, voice memos, meeting recordings, interviews, podcasts, clips, transcripts, transcript search, transcript enrichment, or audio-derived notes with the audio-tools stack.

Zur Installation springen

Repository: learnrudi/registry
Letzte Quellaktivität: 1. August 2026 um 21:27
Erkannte Sprache von SKILL.md: Englisch
Sterne: 0
Forks: 0

Installationsoptionen

Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.

Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.

Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.

npx skills add https://github.com/learnrudi/registry --skill audio-tools-workflow

Der Befehl bleibt in einer Zeile. Scrollen Sie horizontal, um ihn vor dem Kopieren vollständig zu prüfen.

Sie bevorzugen eine lokale Kopie? Laden Sie die Dateien herunter, die SkillsMP derzeit vorliegen.

Quelldateien prüfen

Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.

Datei-Explorer

2 Dateien

SKILL.md

Quellanweisungen · Schreibgeschützte Vorschau

Zurück zur Suche Zurück zur Installation

horizontal-engineering-review

Assess repository-wide architectural coherence across multiple changes, assign evidence-backed dispositions to semantic duplication and shared-contract drift, and plan or execute authorized consolidation. Use for horizontal architecture reviews, repeated mechanisms, consolidation obligations, or cross-module drift; do not use for ordinary single-change code review or Git worktree cleanup.

2026-08-26

map-change-impact

Map the exact repository paths a proposed feature, fix, refactor, migration, or remediation will affect and print the ordered actions, risks, dependencies, and verification needed before implementation. Use when a user asks what file paths will be impacted, what needs to change, what has already been implemented, what remains to do, what the blast radius is, or wants an implementation-ready change map without starting the edits.

2026-08-20

rudi-chief-of-staff

Coordinate complex single- or multi-project objectives from the initiating agent through an acceptance-led task graph, exact provider/model routing, durable run lifecycle, resource and review limits, explicit handoffs, bounded workers, and evidence-backed integration with human oversight. Use when the user asks to delegate dependent work, run a crew, act as a chief of staff or first mate, coordinate multiple agents or desktop tasks, or keep a long-running project moving predictably across worktrees, projects, or hosts.

2026-08-11

Alle aus diesem Repository ansehen

Audio Tools Workflow

Use the local audio-tools MCP stack for speech-to-text workflows. The stack provides audio_transcribe, audio_enrich, audio_sync, audio_stats, and audio_query.

Workflow

Identify the source shape:

local file path: pass file
HTTP(S) direct media or supported video page: pass url
base64 media: pass data and filename

Call audio_transcribe first. It writes JSON and Markdown outputs under ~/.rudi/output/audio-tools/transcripts/YYYY/MM/DD/.

Preserve the returned transcript JSON path. Use it for enrichment and indexing.

Call audio_enrich only when the user needs metadata such as title, summary, tags, topics, people, sentiment, keywords, or action items.

Call audio_sync after transcription or enrichment when query/search/index results matter.

Use audio_stats for a quick collection overview.

Use audio_query for focused analysis across the SQLite transcript index.

Guardrails

Treat transcripts as private user data.

Do not include credentials, API keys, or raw secret values in prompts, outputs, config files, or stack packages.

Only pass HTTP(S) URLs to URL input. The stack rejects other URL schemes.

Prefer read-only SELECT queries with audio_query. Do not run destructive SQL unless the user explicitly asks for database modification.

If enrichment fails because no prompt template is configured, report that transcription completed and name the missing configuration requirement.

If a private source requires authorization, use the host connector to obtain the source file or URL before invoking the stack.

If audio_transcribe reports a missing binary or Whisper model, treat that as a runtime provisioning issue in the MCP/plugin environment. Do not assume a successful manual shell command means the MCP server can see the same dependency.

Useful Files

Connector hints: connectors/host-provided.json

Environment and secret requirements: secrets/required.json

MCP launch contracts: mcp/servers.json

Transcript JSON schema: outputs/schemas/transcript-record.schema.json

Query examples: references/query-examples.md