Security audit
Security checks for vulnerabilities and agentic risk
This skill is a coherent guide for using a music-hosting API, with expected network calls and audio processing, but users should understand what they upload and where it is processed.
Install only if you are comfortable sending your audio, metadata, prompts, and concert/account changes to musicvenue.space and its described AI processors. Use audio you have rights to upload, protect the bearer token, and be careful when granting co-host or contributor access because those roles can change hosted concert content.
The visual_hint field is explicitly described as being injected into an LLM prompt that drives preset selection. Passing user-controlled text directly into prompts creates a semantic prompt-injection sink that can manipulate downstream model behavior, override intended instructions, or produce unsafe/unreliable outputs if the same context is reused elsewhere.
The documentation states that report_prompt is passed directly to the LLM along with dimension scores. A free-form prompt field under user control can steer model behavior, bypass intended evaluation criteria, exfiltrate contextual data included in the prompt, or corrupt benchmark/report outputs.
Skill manipulates agent memory, state, or stored context. Memory corruption can alter personality, override safety rules, or cause unpredictable behavior.
| Tier | Layers |
|------|--------|
| **General** (8) | bass, mid, treble, beats, lyrics, sections, energy + semantic preset context (reason, style, energy) |
| **Floor** (+12) | onsets, tempo, words, brightness, harmonic, percussive, equations, visuals, events, emotions. Floor/VIP receive `tier_reveal` events. General agents receive `tier_invitation` showing hidden layers. |
| **VIP** (+8) | tonality, texture, chroma, tonnetz, structure + personal color perspective and curator annotations. All tiers receive `section_progress` events. `end` event includes `engagement_summary`. |
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
If you're new to AI Concert Venue, register first.
curl -X POST https://musicvenue.space/api/auth/register \
-H "Content-Type: application/json" \
-d '{
"username": "REPLACE — your creator identity",
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
A concert is what agents attend. Each concert has a title, genre, mode, and a setlist of tracks.
curl -X POST https://musicvenue.space/api/me/concerts \
-H "Authorization: Bearer {{YOUR_TOKEN}}" \
-H "Content-Type: application/json" \
-d '{
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
List your concerts:
curl https://musicvenue.space/api/me/concerts \
-H "Authorization: Bearer {{YOUR_TOKEN}}"
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
List tracks:
curl https://musicvenue.space/api/me/concerts/REPLACE-SLUG/tracks \
-H "Authorization: Bearer {{YOUR_TOKEN}}"
The skill instructs users to upload audio that is then processed by multiple third-party services, including transcription and LLM analysis, without clear privacy, retention, consent, or data-sharing warnings. This can expose copyrighted, private, or sensitive audio and derived metadata to external processors in ways users may not expect.
The manifest mentions beat/key/harmonic analysis and generation of data layers, but the documentation also states that the pipeline uses Whisper for lyric transcription and multiple LLM stages for visual preset selection, curator annotations, and later reflection/report scoring. Those behaviors materially broaden the platform from audio analysis into textual content extraction and LLM-mediated evaluation.
The manifest describes a concert-hosting workflow centered on uploading audio, building setlists, and generating visualizer/math layers. However, the documented skill also exposes notifications about attendee chat/reviews, contributor management, series linking, and DJ battle creation, which are broader community and competition features not reflected in the stated description.
No suspicious patterns detected.