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 calesthio/generative-media-skills --skill hume-octavegit clone --depth 1 https://github.com/calesthio/generative-media-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/calesthio/generative-media-skills/hume-octave)<a href="https://agentmods.dev/skills/calesthio/generative-media-skills/hume-octave"><img src="https://agentmods.dev/badge/skills/calesthio/generative-media-skills/hume-octave/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/calesthio/generative-media-skills/hume-octave"><img src="https://agentmods.dev/badge/skills/calesthio/generative-media-skills/hume-octave.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 6 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Data Exfiltration · line 201 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Data Exfiltration · line 220 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Data Exfiltration · line 239 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Data Exfiltration · line 267 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Data Exfiltration · line 305 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Excessive Agency · line 313 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00059 | $0.06013 |
| Opus 5 | $0.00030 | $0.03007 |
| Sonnet 5 | $0.00012 | $0.01203 |
| Haiku 4.5 | $0.00006 | $0.00601 |
Grade A, and why
hume-octave scanned grade A with 1 finding 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 9d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl https://api.hume.ai/v0/tts \ How it starts
The opening of the file, as written. The whole thing — 387 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Hume Octave production guidance
Use Hume Octave when the speech performance matters as much as the words: character narration, emotionally varied voiceover, interactive playback, branded voices, multilingual narration, voice cloning with consent, voice conversion, or audio that needs word/phoneme timestamps.
Do not treat Octave as a generic flat TTS engine. It is strongest when the request needs semantic context, emotional delivery, voice identity, or continuity across related utterances.
Do not use Octave for music, singing, non-speech sound effects, full audio mixing/mastering, source separation, or video generation. Use a separate audio post tool for loudness, noise cleanup, ducking, fades, file normalization, and final mix delivery.
Verification note: provider facts in this skill were checked against official Hume documentation, API reference, pricing, changelog, terms, privacy, and API data-usage pages on 2026-07-10. Re-check volatile facts before quoting exact pricing, plan limits, language lists, model status, or retention controls.
What Octave can and cannot do
Documented facts:
- Octave is Hume's speech-language model behind its TTS and voice features. Hume describes it as using language-model intelligence to understand text semantically and emotionally, rather than only reading phonemes. Source: https://dev.hume.ai/docs/text-to-speech-tts/overview
- The TTS API supports streaming HTTP output, WebSocket bidirectional streaming, and non-streaming HTTP responses. Source: https://dev.hume.ai/docs/text-to-speech-tts/overview
- Main TTS endpoints are
/v0/tts,/v0/tts/file,/v0/tts/stream/json,/v0/tts/stream/file, and WebSocketwss://api.hume.ai/v0/tts/stream/input. Source: https://dev.hume.ai/reference/text-to-speech-tts/synthesize-json and https://dev.hume.ai/reference/text-to-speech-tts/stream-input - Octave 1 supports English and Spanish. Octave 2 preview supports English, Japanese, Korean, Spanish, French, Portuguese, Italian, German, Russian, Hindi, and Arabic. Source: https://dev.hume.ai/docs/text-to-speech-tts/overview
- Hume documents model latency around
~200msfor Octave 1 and~100msfor Octave 2 preview, excluding network transit. Streaming instant mode typically has first audio ready around~200ms, depending on load and input complexity. Source: https://dev.hume.ai/docs/text-to-speech-tts/overview - The TTS API limit is 5,000 characters per utterance, 1,000 characters per utterance description, up to 5 generations per request, and output formats
MP3,WAV, andPCM. Source: https://dev.hume.ai/docs/text-to-speech-tts/overview - An utterance can include
text, optionalvoice, optionaldescription,speed, andtrailing_silence. Source: https://dev.hume.ai/reference/text-to-speech-tts/synthesize-json speedis a non-linear scale from0.5to2.0;1.0is normal. Source: https://dev.hume.ai/docs/text-to-speech-tts/acting-instructionsdescriptionacting instructions are documented for Octave 1 only; Hume says Octave 2 support is coming soon.speedandtrailing_silenceare supported in all models. Source: https://dev.hume.ai/docs/text-to-speech-tts/acting-instructions- If you specify
version: "2", Hume requires avoice; dynamic voice generation without a predefined voice is rejected for Octave 2. Source: https://dev.hume.ai/reference/text-to-speech-tts/synthesize-json - Voice design from natural-language descriptions currently requires Octave 1, and voices designed with Octave 1 are compatible with Octave 2 requests. Hume says multilingual voice design for Octave 2 is coming soon. Source: https://dev.hume.ai/docs/voice/voice-design
- Voices may be referenced by
idor byname; when using a name, includeprovider: "HUME_AI"for Hume Voice Library voices orprovider: "CUSTOM_VOICE"for saved custom voices. Source: https://dev.hume.ai/docs/text-to-speech-tts/voice - Voice cloning can be created from a microphone recording or uploaded audio sample from a consenting speaker; access depends on subscription tier. Hume's TTS overview says cloning can work with as little as 15 seconds of audio. Source: https://dev.hume.ai/docs/voice/voice-cloning and https://dev.hume.ai/docs/text-to-speech-tts/overview
- Voice conversion accepts speech audio plus a target voice and returns converted audio while preserving the source speech patterns, timing, and emotional expression. Source: https://dev.hume.ai/docs/text-to-speech-tts/voice-conversion
- Voice conversion input audio should be human speech, clear, at least 12 seconds and less than 3 minutes, with supported input formats
MP3,WAV,M4A, andOGG; 44.1 kHz is recommended. Source: https://dev.hume.ai/docs/text-to-speech-tts/voice-conversion - Octave 2 can return word- and phoneme-level timestamps when requested via
include_timestamp_types; timestamp requests requireversion: "2". Source: https://dev.hume.ai/docs/text-to-speech-tts/timestamps - Phoneme-level timestamps use IPA symbols, with IPA-compatible extensions for some languages. Source: https://dev.hume.ai/docs/text-to-speech-tts/timestamps
- Pricing, monthly included TTS characters, request-per-minute limits, voice cloning availability, and commercial-use permissions vary by plan. Source: https://www.hume.ai/pricing and https://www.hume.ai/terms-of-use
- Hume's terms say Free and Starter plans are limited to non-commercial use, while Creator and above may be used commercially subject to Hume policies. Source: https://www.hume.ai/terms-of-use
- Hume's API Data Usage Policy says customer-submitted API data is not used to train Hume models or improve service offerings. The Terms of Use and Privacy pages should still be checked for license, retention, account settings, and non-API Playground behavior. Sources: https://www.hume.ai/api-data-usage-policy, https://www.hume.ai/terms-of-use, https://dev.hume.ai/docs/resources/privacy
What ships with it
1 file 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.
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.
- 9d ago First seen · 387 lines · 59 tokens per session scan A c1e02c532612
hume-octave is a skill published in the GitHub repository calesthio/generative-media-skills (171 stars, last pushed 2mo ago), licensed MIT. It adds 59 tokens to every session and 6,013 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
cliptalk-cover-director
Produces evidence-backed cover candidates and reviewable cover variants for a ClipTalk video. Use when the user asks for a cover, poster frame, thumbnail, or multiple cover directions; do not use for timeline editing or social-video reframing.
cliptalk-smart-reframe
Creates a subject-aware, time-varying crop track and a review-only social-format preview from an accepted ClipTalk cut. Use for automatic vertical, square, or portrait reframing; do not use for a fixed manual crop or before content editing is accepted.
cliptalk-content-extractor
Locates and assembles source passages matching a semantic request. Use for extracting explanations, topics, quotes, demonstrations, or other specifically described content.
cliptalk-interview-editor
Produces a coherent interview edit by combining speaker discovery, topic selection, dialogue context, cleanup, subtitles, and preview. Use for interviews, podcasts, testimonials, or question-and-answer recordings.
cliptalk-shortform-hook-director
Finds and assembles a reviewable short-form cut with a strong opening hook. Use for Shorts, Reels, social clips, talking-head cutdowns, or requests for a punchier opening.
cliptalk-social-reframe-exporter
Creates a review-only 9:16, 4:5, 1:1, or 16:9 version from an existing accepted ClipTalk cut, then checks the rendered preview. Use only when a cut already exists and the user asks to adapt it for Shorts, Reels, Douyin, Xiaohongshu, WeChat Channels, or square feeds; do not use when the user still needs content found…