Lunara
Server Details
Subtitle engine for AI agents: one upload gives a written track and a translation, timings exact.
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-06-18
- URL
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 22 tools
Most tools target a distinct stage of the subtitle pipeline (probe, transcribe, rewrite, translate, QA, coverage, readability), and the guidance tools (guide, list_tools, next_step, whoami) are separated by audience and purpose. The one real collision is lunara_start_subtitle_job vs lunara_subtitle_from_url, which both start a full pipeline job from an already-hosted URL; the descriptions say 'prefer' each other in almost identical terms, so an agent can easily pick either. Glossary_get vs glossary_extract is also mild overlap but the read-vs-discover framing clarifies it.
All names share the lunara_ prefix and snake_case, with a dominant verb_noun pattern (get_job, list_cues, start_subtitle_job, upload_video, probe_media). Deviations exist — noun_verb (glossary_extract), noun-only (guide, readability, whoami), and noun phrases (subtitle_from_url, next_step, qa_contract) — but the prefix and casing keep the set readable and predictable.
22 tools sits in the heavy band for a single-domain media/subtitle service. Four are pure navigation/help tools (guide, list_tools, next_step, whoami) and several readout tools (coverage, readability, qa_contract, list_cues) could plausibly be folded together, though each does answer a distinct question.
The core lifecycle is covered end to end: upload → finish_upload → job polling → SRT, plus transcribe/rewrite/translate/segment and quality gates. However there are dead ends: list_cues explicitly supports 'proposing edits' yet no tool applies a cue edit, glossary_extract surfaces candidates but nothing writes them to the glossary, and there is no job cancel/delete or job-history listing.
Available Tools
22 toolslunara_finish_uploadAIdempotentInspect
Tell Lunara the bytes are uploaded so the pipeline can start. Returns the job id to poll with lunara_get_job.
Inputs: key (required), parts (required), videoId (required), uploadId (required), +1 more.
lunara_get_coverageARead-onlyIdempotentInspect
Report how much of the source timeline the subtitles cover, and list suspicious gaps longer than a second. Use before shipping a long video.
Inputs: jobId (required).
lunara_get_glossaryARead-onlyIdempotentInspect
Return the proper nouns Lunara has learned for this org, so names are spelled consistently across tracks.
Inputs: limit.
lunara_get_jobARead-onlyIdempotentInspect
Return the state of a subtitle job: status, progress, whether the SRT is ready, and how many cues it has. Poll this after lunara_upload_video.
Inputs: jobId (required).
lunara_get_queueARead-onlyIdempotentInspect
How many transcription jobs are running and whether a new one can be accepted. Call this before lunara_start_subtitle_job or lunara_subtitle_from_url if you are submitting several jobs — when saturated, wait the retryAfterSeconds rather than retrying immediately.
Inputs: no named parameters.
lunara_get_srtARead-onlyIdempotentInspect
Return the subtitle file for a finished job. One line per cue. Use track="secondary" for the translated track (English by default).
Inputs: jobId (required), track.
lunara_glossary_extractARead-onlyIdempotentInspect
Find candidate proper nouns in a job’s cues, ranked by frequency, with the first cue they appear in. Suggestions for a glossary, not corrections — nothing is rewritten.
Inputs: jobId (required), limit.
lunara_guideARead-onlyIdempotentInspect
Returns a step-by-step guide written to be read by a human, not by you. Call this when someone asks "how do I use this", "what can it do", "how do I get started", or when you need to explain the workflow rather than execute it. Pass journey="first_subtitle" for the end-to-end path, "api_key" for credential setup, "languages" for what is supported, or "limits" for what it cannot do. Answer the person with this text rather than paraphrasing from memory.
Inputs: journey.
lunara_list_cuesARead-onlyIdempotentInspect
Return subtitle cues for a job as timestamps plus text, so you can inspect or reference individual lines before proposing edits.
Inputs: jobId (required), limit, offset.
lunara_list_toolsARead-onlyIdempotentInspect
Return every subtitle tool Lunara exposes, with the minutes each one costs. Call this first when you are unsure which tool fits.
Inputs: no named parameters.
lunara_next_stepARead-onlyIdempotentInspect
Tell me what to call next. Given a jobId (or nothing), returns ready-to-paste tool calls with the exact arguments, plus one line of why. Use this when you are unsure of the order — or just read the "next" array that every mutating tool already returns. Free.
Inputs: jobId.
lunara_probe_mediaARead-onlyIdempotentInspect
Return duration, dimensions, codecs and container for a URL, plus whether the audio is mono 16 kHz already. Call this before choosing an ASR engine.
Inputs: mediaUrl (required).
lunara_qa_contractARead-onlyIdempotentInspect
Run the contract gate over both tracks: ASR text and timings byte-exact, cue length bounds, no empty cues. Returns pass/fail per rule. Free, because refusing a broken track should never cost the caller money.
Inputs: jobId (required), track.
lunara_readabilityBRead-onlyIdempotentInspect
Per-cue reading speed, displayed length and characters-per-second, with the cues that exceed comfortable limits. Mirrors the metrics shown in the editor.
Inputs: jobId (required), track.
lunara_rewriteAIdempotentInspect
Turn spoken Cantonese cue text into written Chinese. Every cue boundary and timestamp is preserved. Deliberate mispronunciations are never corrected, and proper nouns are protected.
Inputs: jobId (required), track.
lunara_segmentAIdempotentInspect
Approximate split, not audio-grounded segmentation: cue text is divided at clause punctuation and each piece receives a share of the original span proportional to its character count. The words are not changed. Timings are estimated, so treat them as a reading aid rather than a re-alignment — use lunara_start_subtitle_job if you need timing derived from the audio.
Inputs: jobId (required), track, maxCueMs.
lunara_start_subtitle_jobAInspect
Run the full eight-layer subtitle pipeline on an audio or video URL you can already host. Produces a written-language track and a translated track. Prefer lunara_upload_video when the file is local.
Inputs: audioUrl (required), spokenLang, primaryLang, secondaryLang, +1 more.
lunara_subtitle_from_urlAInspect
Start a persisted subtitle job from a URL you already host — no upload needed. Returns a job id; poll lunara_get_job every 15–30 s until status is "completed", then call lunara_get_srt. Prefer this over lunara_upload_video whenever the media is reachable by URL.
Inputs: audioUrl (required), filename, spokenLang, primaryLang, +2 more.
lunara_transcribeAIdempotentInspect
Run speech recognition only — no rewrite, no translation. Returns timed segments. Use this when you want the transcript itself, or when you plan to pick a later stage yourself.
Inputs: audioUrl (required), spokenLang.
lunara_translateAIdempotentInspect
Translate the text of an existing job’s cues into another language. Every cue boundary and timestamp is preserved; only the text changes. Use this instead of translating yourself so the timing contract holds.
Inputs: jobId (required), targetLang (required), track.
lunara_upload_videoAInspect
Reserve an upload slot and return the object key plus a signed URL per part. After uploading the bytes call lunara_finish_upload. Use this for a file you hold locally.
Inputs: mime (required), filename (required), sizeBytes (required), projectId, +6 more.
lunara_whoamiARead-onlyIdempotentInspect
Return the organisation, user, plan and remaining balance behind the credential you are using. Call this FIRST when someone is unsure which account will be charged, or after they say the work went to the wrong account. Every other tool bills the organisation reported here.
Inputs: no named parameters.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Generate SRT and VTT subtitles for any video or audio file, 99 languages.
Transcribe audio and video, and translate subtitles into 50+ languages.
YouTube transcripts for AI agents: captions with timestamps, languages, TXT, SRT and VTT export.
YouTube transcripts for AI agents: 150+ languages, AI transcription fallback, search, summaries.
Related MCP Servers
- AlicenseAqualityBmaintenanceEnables an AI agent to inspect, repair, retime, convert, merge/split, and search SRT, WebVTT, and ASS subtitle files entirely offline, with exact-fraction resync, dry-run fixes, and honest encoding detection. Translation is handled by the agent itself through a chunk/apply contract that preserves timing and flags lost tags.10MIT
- AlicenseAqualityBmaintenanceEnables AI agents to convert any-length text to speech in saved or cloned voices, generate SRT/VTT captions from local media, and produce per-scene narration with exact durations and frame timings for video timelines.5MIT
- AlicenseAqualityBmaintenanceEnables AI agents to turn scripts into narration audio with matching SRT/VTT captions in one tool call, plus voice cloning from a reference clip and transcription of local audio or video into subtitles.7MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to transcribe and summarize videos and voice messages, translate speech while keeping the speaker's voice, read text aloud in a cloned voice, and produce subtitles, lip-synced dubbing, voice changes, and translated images from links, uploads, or local files. It also handles media downloading and job/account status, with paid operations priced and confirmed before they run.MIT
Glama MCP Gateway
Add one secure layer between your agents and this server.