Skillv1.0.1

ClawScan security

preftrade AIO Quantitative Research Tools · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousApr 27, 2026, 2:44 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The onboarding instructions mostly match the stated purpose (onboarding an MCP client), but there are small coherence issues — notably the metadata declares no required environment variables while the runtime instructions rely on a PREFERENCE_API_KEY and on writing a credentials file to ~/.config — so verify trust in the external service before installing.
Guidance: This skill appears to be an onboarding guide for pref.trade's MCP and is mostly coherent, but check these before installing: - Trust the service: confirm you trust https://pref.trade (review privacy, terms, and reputation). This skill will get an agent API key that grants requests against their platform. - Metadata mismatch: the registry declares no required env vars but the instructions use PREFERENCE_API_KEY — expect to provide that env var or a credentials file. Ask the publisher to update metadata for clarity. - Secret storage: the guide recommends saving the key to ~/.config/preference/credentials.json or an env var. Storing secrets on disk is common but increases on-disk persistence risk — ensure file permissions are correct (chmod 600) and rotate/delete keys when no longer needed. - Avoid accidental leaks: follow the skill's own warnings — do not paste the key into chat, logs, or public places and avoid running commands that print the key into shell history. Prefer using a short-lived or limited-scope key if the service supports it. - Test conservatively: consider using anonymous or low-quota testing first and verify behavior with minimal privileges before registering high-privilege account keys. If you want a stronger assurance, ask the skill author to: (1) add PREFERENCE_API_KEY to the declared required env vars in metadata, (2) document exact scopes/permissions of pref_agent_* keys, and (3) provide a link to the service's security/privacy policy. If those are missing, treat the metadata inconsistency as a red flag.

Review Dimensions

Purpose & Capability: okThe skill's name and description describe onboarding an MCP (pref.trade) and the instructions focus on registering an agent key, storing it, and configuring an MCP client — all consistent with a pref.trade onboarding tool.
Instruction Scope: noteInstructions tell the agent to call pref.trade endpoints, create/save an agent API key, and configure MCP clients. This stays within the stated purpose. Note: it instructs writing a secret to ~/.config/preference/credentials.json (expected for client config) and suggests running debug commands like echo $PREFERENCE_API_KEY which can expose secrets if used carelessly. The guidance includes warnings not to echo keys into history/logs.
Install Mechanism: okThere is no install spec and no code to download or execute; the skill is instruction-only, which minimizes on-disk risk.
Credentials: concernThe registry metadata lists no required environment variables, but the SKILL.md repeatedly references PREFERENCE_API_KEY and gives commands that rely on that env var or on a credentials file. That mismatch is an incoherence in metadata vs runtime expectations. Aside from that, the only secret involved is the pref_agent_* key — proportionate to the described function — but storing it on disk (in ~/.config) is recommended by the skill and has the usual persistence risks.
Persistence & Privilege: okThe skill does not request 'always' or elevated privileges and does not modify other skills or system-wide settings. It asks the user/agent to create a credentials file in the user's config dir — a normal client configuration step.