Security audit
Security checks for vulnerabilities and agentic risk
This is a straightforward Lidarr control skill that can change a media library and trigger searches, but its behavior is visible and purpose-aligned.
Install only if you want an agent to operate your Lidarr instance. Treat add, search, delete, blacklist, and queue commands as real changes that may start downloads or remove items; confirm targets before running them and keep the Lidarr API key in a protected secret file.
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
# 1. Look up artist (returns MusicBrainz results)
curl -s "$LIDARR_URL/api/v1/artist/lookup?term=David+Bowie" \
-H "X-Api-Key: $LIDARR_KEY" | python3 -c "
import sys,json
results = json.load(sys.stdin)
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
"
# 2. Get quality/metadata profiles and root folders
curl -s "$LIDARR_URL/api/v1/qualityprofile" -H "X-Api-Key: $LIDARR_KEY"
curl -s "$LIDARR_URL/api/v1/metadataprofile" -H "X-Api-Key: $LIDARR_KEY"
curl -s "$LIDARR_URL/api/v1/rootfolder" -H "X-Api-Key: $LIDARR_KEY"
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
"
# 2. Add the album (artist must already be in Lidarr)
curl -s -X POST "$LIDARR_URL/api/v1/album" \
-H "X-Api-Key: $LIDARR_KEY" \
-H "Content-Type: application/json" \
-d '{"foreignAlbumId":"<id>","monitored":true,"addOptions":{"searchForNewAlbum":true}}'
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
curl -s -X POST "$LIDARR_URL/api/v1/command" \
-H "X-Api-Key: $LIDARR_KEY" \
-H "Content-Type: application/json" \
-d '{"name":"MissingAlbumSearch"}'
The skill documents state-changing Lidarr operations such as adding artists/albums and triggering searches/download workflows without any explicit warning, confirmation step, or guidance to verify user intent first. In an agent context, this increases the chance of unintended modifications to the media library and unnecessary downloader activity, even though the commands target a local service.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
"
# 2. Get quality/metadata profiles and root folders
curl -s "$LIDARR_URL/api/v1/qualityprofile" -H "X-Api-Key: $LIDARR_KEY"
curl -s "$LIDARR_URL/api/v1/metadataprofile" -H "X-Api-Key: $LIDARR_KEY"
curl -s "$LIDARR_URL/api/v1/rootfolder" -H "X-Api-Key: $LIDARR_KEY"
This POST request adds an album and can trigger a new album search, causing state changes and potentially downstream activity in download clients. In the context of an autonomous agent skill, unaudited write operations to a local media-management API are security-relevant because they can alter configuration-managed content and initiate unwanted tasks.
"
# 2. Add the album (artist must already be in Lidarr)
curl -s -X POST "$LIDARR_URL/api/v1/album" \
-H "X-Api-Key: $LIDARR_KEY" \
-H "Content-Type: application/json" \
-d '{"foreignAlbumId":"<id>","monitored":true,"addOptions":{"searchForNewAlbum":true}}'
This command triggers a global MissingAlbumSearch, which can initiate broad downloader activity and consume network, indexer, and client resources. While intended for legitimate automation, exposing it without safety guidance or confirmation makes unintended bulk actions more likely in an agent-driven environment.
curl -s -X POST "$LIDARR_URL/api/v1/command" \
-H "X-Api-Key: $LIDARR_KEY" \
-H "Content-Type: application/json" \
-d '{"name":"MissingAlbumSearch"}'
The API reference documents destructive operations such as deleting artists, albums, track files, and queue items including blacklisting, but provides no warning about irreversible effects or operational safeguards. In a skill designed to drive an agent's actions, this increases the chance of accidental destructive requests being issued from natural-language prompts or automation mistakes.
The document specifies use of an API key header but gives no guidance on secure handling, storage, redaction, or avoiding exposure in logs and prompts. In an agent skill context, credentials may be propagated through tool traces or debugging output, creating avoidable secret-leak risk.
No suspicious patterns detected.