Wavye
SuspiciousAudited by ClawScan on May 10, 2026.
Overview
Wayve appears purpose-aligned for planning and coaching, but it asks the agent to persist broad personal, health, schedule, and business details to Wayve and to mutate the user's Wayve account through an unpinned CLI.
Install only if you want Wayve to act as a persistent planning and coaching memory. Before using it, understand what will be saved to Wayve, avoid sharing highly sensitive details unless you want them stored, keep your WAYVE_API_KEY secure, review automations and knowledge entries regularly, and prefer a pinned/verified CLI version if possible.
Findings (5)
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.
Personal and business details shared during planning could be stored long-term and reused later, including sensitive health, family, revenue, and crisis context.
The artifact explicitly describes a persistent, server-side memory system that may store sensitive health, family, financial, business, and crisis information for reuse across sessions and clients.
The knowledge base is stored server-side via the Wayve API, so it persists across devices, sessions, and even different AI clients... `health_conditions`... `family_situation`... `revenue_monthly`... `crisis_event_2026_03`
Only use this if you are comfortable with Wayve storing persistent memory. Ask the agent before saving sensitive details, review/delete knowledge entries regularly, and avoid sharing information you do not want stored server-side.
The agent can change your Wayve knowledge base, activities, audits, automations, and settings when using the skill.
The skill directs the agent to run account-mutating CLI commands through Bash. This is central to Wayve's purpose, but it can create, update, or delete user data.
Use the Bash tool with the `wayve` command... `wayve knowledge summary/list/save/update/delete --json`... `wayve activities create/update --json`... `wayve settings get/update --json`
Confirm important changes, especially deletes, bulk updates, automation creation, and settings changes. Review the CLI output before relying on the result.
Anyone with the API key may be able to access or modify Wayve account data depending on the key's permissions.
The skill requires an API key that likely grants access to the user's Wayve account and data. This is expected for the integration but is still a meaningful credential boundary.
Required env vars: WAYVE_API_KEY... Primary credential: WAYVE_API_KEY
Use a dedicated, revocable Wayve key if available, store it securely, rotate it if exposed, and avoid sharing unrelated credentials.
Future installs may run a different CLI version than the one intended when the skill was published.
The runtime dependency is installed from npm using the mutable `@latest` tag, and no CLI source code was included for review. This is disclosed and purpose-aligned, but it means the installed code can change after review.
node | package: @gowayve/wayve-cli@latest | creates binaries: wayve
Prefer a pinned, reviewed CLI version and verify the package source before installing, especially because the CLI handles credentials and personal data.
If approved, Wayve automations may continue running or producing stored analysis after the current conversation.
The skill supports scheduled/background analyses that can save results later. The artifact requires explicit confirmation and one-time defaults, which mitigates the risk, but users should understand these jobs may persist until disabled or deleted.
propose to the user to schedule them at night... create an agent routine... After the analysis runs, disable or delete the automation... Never create background jobs without explicit user confirmation.
Approve background jobs only when you want them, keep them specific and one-time by default, and periodically list/delete automations you no longer need.