Slybroadcast Voicemail
WarnAudited by ClawScan on May 10, 2026.
Overview
This skill is clearly meant to send Slybroadcast voicemail campaigns, but it gives an agent high-impact account and campaign authority without enough scoping, credential disclosure, or reviewed implementation details.
Review this skill carefully before use. Only connect a trusted, reviewed implementation; use limited credentials; require explicit approval before every send or campaign-control action; confirm recipient consent and legal/compliance requirements; and manage public audio staging so voicemail content is not left exposed.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
An agent could send voicemail campaigns or control existing campaigns on the user's account, potentially reaching many phone numbers and incurring account, reputational, or compliance impact.
The skill gives the agent tools to send and control voicemail campaigns, including bulk-style sends, but does not define confirmation, consent, cost, rate-limit, or recipient-scope safeguards.
Use this skill when the user wants to send one or many voicemail drops... Tool names: - `slybroadcast_voicemail_send` ... - `slybroadcast_campaign_control`
Require explicit per-campaign user confirmation, verify recipients and consent, use small test sends first, and avoid allowing autonomous sends or campaign controls.
Users may not realize before install that the skill needs credentials capable of using paid third-party services and sending outbound communications.
The skill requires account passwords/API keys for services that can send campaigns or generate voice audio, but the registry metadata declares no required env vars or primary credential.
Required environment variables: - `SLYBROADCAST_UID` ... - `SLYBROADCAST_PASSWORD` ... For ElevenLabs voice generation: - `ELEVENLABS_API_KEY`
Declare all credentials in metadata, use least-privilege service credentials where possible, avoid sharing personal account passwords, and keep campaign-sending credentials separate from other accounts.
The actual code that would handle credentials and send/control campaigns is not available in the artifacts, so users cannot verify what would run.
The reviewed package contains only SKILL.md and no install spec or code, yet it instructs use of an npm workspace and MCP server that are not present for review.
npm --workspace @fub/slybroadcast-voicemail run dev:cli -- send --help ... npm --workspace @fub/slybroadcast-voicemail run dev:mcp
Only use this with a reviewed, pinned, trusted implementation; provide the package source, install instructions, lockfile, and MCP/CLI code before granting credentials.
Voicemail audio or AI-generated message content may be accessible through a public URL if staging is not protected or cleaned up.
The skill discloses that generated or local audio must be placed at a publicly reachable location, which is purpose-aligned but creates an exposure point for message content.
Local and AI-generated files are staged first and must be publicly reachable for Slybroadcast to fetch.
Do not include sensitive content in staged audio, use hard-to-guess temporary URLs if supported, restrict access where possible, and delete staged files after Slybroadcast has fetched them.