Skillv1.0.0

ClawScan security

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

BenignFeb 28, 2026, 10:49 AM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: This is an instruction-only personal coaching skill whose declared purpose matches its runtime instructions and it does not request credentials or install code, but you should confirm how (and whether) it will integrate with other tools before enabling it broadly.
Guidance: This skill is low-risk: it is instruction-only, requests no credentials, and does not install code. Before enabling it broadly, ask the publisher (or inspect the runtime integration) to clarify: (1) what the listed integrations (habit-tracker, task-manager, daily-digest, personality-modes) actually access and whether they require tokens or data access, and (2) whether the COACH_* environment variables will be read by the agent at runtime and need to be set. If you plan to let the agent invoke the skill autonomously, consider limiting its scope or monitoring its activity initially (check logs, watch for any prompts that ask for account passwords or API keys). If the skill later requests access to external services or credentials, require explicit review before granting those secrets.

Purpose & Capability: okName and description (AI coaching, goals, check-ins, motivation) match the SKILL.md content and example commands; the behavior described is consistent with a conversational coach.
Instruction Scope: okSKILL.md contains only conversational flows, example commands, suggested env config names, and high-level mentions of integrations. It does not instruct the agent to read local files, exfiltrate data, call arbitrary external endpoints, or perform unrelated privileged actions.
Install Mechanism: okNo install spec and no code files — this is instruction-only, so nothing is written to disk or downloaded during install.
Credentials: noteSKILL.md shows optional COACH_* environment variables for configuration, but the skill metadata declares no required env vars. This is a minor mismatch (the SKILL.md suggests configurable env vars but doesn't demand secrets or other credentials). No sensitive credentials are requested.
Persistence & Privilege: okalways is false and the skill does not request persistent system-wide privileges or modify other skills' configs. Autonomous model invocation is allowed (platform default) but not elevated here.