Back to skill
Skillv1.1.0
ClawScan security
Sleep Channel · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignApr 4, 2026, 11:25 PM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill does what it claims (a deterministic Discord sleep-channel tracker) and requires only local file access and a timezone setting; there are minor documentation inconsistencies but no signs of exfiltration or unrelated privileges.
- Guidance
- This skill appears to be what it says: a local, deterministic sleep tracker that parses Discord messages and writes logs under your workspace. Before installing: 1) Note the SKILL.md requires SLEEP_TIMEZONE (IANA timezone) but the registry metadata omitted that — set SLEEP_TIMEZONE and optionally WORKSPACE. 2) The tracker will create/modify files at $WORKSPACE/data/sleep_log.csv and $WORKSPACE/state/sleep_summary_state.json — ensure that location is secure and you are comfortable storing message IDs/author IDs there. 3) There are no network calls or secret credentials requested, so it won’t exfiltrate data by design, but treat the workspace files as sensitive. 4) Prefer running the skill in a dedicated agent/channel and restrict file permissions (e.g., use a per-skill workspace or container) if you want stronger isolation. 5) If you need absolute assurance, inspect the included scripts/tracker.py (already provided) and test in an isolated environment first. If you want me to, I can produce a shortened checklist or suggest a safe container invocation for running this skill.
Review Dimensions
- Purpose & Capability
- noteThe name/description align with the included script and instructions: parsing Discord messages and logging sleep events to local CSV/JSON files. Minor incoherence: registry metadata lists no required environment variables, but SKILL.md and the script require SLEEP_TIMEZONE (and optionally WORKSPACE). This is likely a packaging/documentation omission rather than malicious.
- Instruction Scope
- okSKILL.md keeps scope narrowly to calling scripts/tracker.py for add/correct/delete/render actions and mandates using source Discord timestamps. It does not instruct reading unrelated files, contacting external endpoints, or accessing secrets beyond environment config. It does rely on the agent providing Discord metadata (channel_id, message_id, author_id, timestamp_utc).
- Install Mechanism
- okNo install spec; the skill is instruction-plus-a-local-Python-script using only the stdlib. No network downloads or external packages — low install risk.
- Credentials
- noteThe script requires SLEEP_TIMEZONE (declared required in SKILL.md) and optionally WORKSPACE; the code will default to UTC/~/ .openclaw/workspace if not set. It does not request secrets or external credentials. Important: the tracker writes message IDs, author IDs, timestamps and notes to data/sleep_log.csv and state/sleep_summary_state.json under WORKSPACE — users should be aware of the privacy implications and file permissions. The earlier registry metadata omitted the required env var, which is an inconsistency to fix.
- Persistence & Privilege
- okalways:false and user-invocable:true (normal). The skill writes its own data files under the workspace and manages its own state file; it does not request system-wide changes or modify other skills.