Skillv1.0.3

ClawScan security

Authbridge · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignApr 22, 2026, 7:16 PM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's instructions, requirements, and behavior are internally consistent with an Authbridge integration that uses the Membrane CLI; nothing requested is disproportionate to the stated purpose.
Guidance: This skill is coherent and appears to do what it says: it uses the Membrane CLI to manage Authbridge connections and actions. Before installing, verify you trust the @membranehq/cli package and its publisher (check the npm package page and the GitHub repo referenced), prefer a local or non-global install if you want to limit system-wide changes, and run CLI commands from a non-sensitive environment until you confirm behavior. Understand that network access and browser-based authentication are required and that you are delegating credential management to Membrane, so only proceed if you’re comfortable trusting their service.

Purpose & Capability: okThe name/description (Authbridge integration) matches the instructions: the SKILL.md tells the agent to use the Membrane CLI to create a connection, discover and run actions against Authbridge. No unrelated credentials, binaries, or system paths are requested.
Instruction Scope: okRuntime instructions are scoped to installing and using the Membrane CLI (login, connect, action list/create/run). They do not instruct reading arbitrary local files, accessing unrelated env vars, or exfiltrating data to unexpected endpoints. The guidance to use browser-based auth and headless flow is explicit.
Install Mechanism: noteThere is no bundled install spec, but the docs recommend installing @membranehq/cli via npm install -g. Using a public npm package is expected for a CLI integration; this carries the usual moderate risk of trusting an npm package and running global installs. No downloads from untrusted URLs or archives are suggested.
Credentials: okThe skill declares no required environment variables or config paths and explicitly advises against asking for API keys. Authentication is delegated to Membrane, which aligns with the skill's purpose and removes the need for local secrets.
Persistence & Privilege: okThe skill is not forced always-on (always: false) and does not request modification of other skills or system-wide settings. Autonomous invocation is allowed (platform default) but not combined with other concerning flags.