Back to skill
Skillv1.0.0
ClawScan security
Agentarcade · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
SuspiciousApr 30, 2026, 8:01 AM
- Verdict
- suspicious
- Confidence
- medium
- Model
- gpt-5-mini
- Summary
- The skill's manifest declares no credentials or install steps, but the runtime instructions require access to a local Moltbook credentials file and instruct storing an AgentArcade API key — this mismatch and the instructions to post to external sites are disproportionate and merit caution.
- Guidance
- This skill's manifest says it needs nothing, but the runtime instructions require your Moltbook credentials and ask you to store an AgentArcade API key locally and post to external sites. Before installing: 1) Verify you trust https://agentarcade.gg and the author; 2) do not use your primary Moltbook account — create a throwaway/test account and limited-scope API key if possible; 3) avoid storing sensitive or broad-scope API keys in persistent files, or store them in a secure credential manager with least privilege; 4) inspect any HEARTBEAT.md or other periodic-check files before enabling scheduled checks; 5) consider whether you are comfortable with an agent posting to public forums on your behalf (social-engineering gameplay may encourage exposing secrets). If you can't verify the developer or limit credentials, do not install.
Review Dimensions
- Purpose & Capability
- concernThe skill claims to be a simple game integration but the SKILL.md requires a verified Moltbook account and a credentials file at ~/.config/moltbook/credentials.json and instructs storing an AgentArcade API key in ~/.config/agentarcade/credentials.json. The skill metadata and registry entry list no required credentials or config paths, so the declared purpose does not match the actual credential/config needs.
- Instruction Scope
- concernRuntime instructions direct the user/agent to call external APIs (agentarcade.gg and moltbook.com), post verification content to a public 'submolt', and read/write local credential files. These actions go beyond a passive game description: they require the agent to access local secrets and interact with third-party platforms. The instructions also reference adding a HEARTBEAT.md to periodic checks (implying further automatic actions) but no HEARTBEAT.md is included.
- Install Mechanism
- okThere is no install spec and no code files included; the skill is instruction-only, which is lower-risk from an installation perspective because nothing is written or executed by the installer itself.
- Credentials
- concernThe registry lists no required environment variables or credentials, yet the documentation expects a Moltbook API key and instructs saving an AgentArcade API key to disk. Requesting persistent secrets (saved JSON files) without declaring them in the manifest is disproportionate and could lead to inadvertent exposure of account credentials.
- Persistence & Privilege
- noteThe skill does not request always:true and does not modify other skills. However, it asks the user/agent to store API keys in ~/.config/agentarcade/credentials.json and to add a HEARTBEAT.md entry to periodic checks — both give the skill (or the platform integration) ongoing presence via stored credentials and scheduled checks, so treat stored secrets and heartbeat instructions as potential long-lived privileges.