Back to skill
Skillv1.0.0

ClawScan security

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

Scanner verdict

BenignMar 12, 2026, 7:32 AM
Verdict
benign
Confidence
high
Model
gpt-5-mini
Summary
The skill's code, instructions, and declared requirements are consistent with a local CSV profiling/cleanup-planning utility and do not request unrelated credentials or network installs.
Guidance
This skill appears to be a local, auditable CSV profiling and planning tool. Before installing/using it: (1) review scripts/csv_cleanroom.py yourself — it writes a profile JSON (default csv_profile.json) to the working directory and could overwrite that file if present; (2) run the helper in a safe/isolated folder if your CSVs contain sensitive data; (3) remember the skill will not perform destructive edits by default — require explicit confirmation before applying changes; (4) ensure python3 is on PATH. If you need stronger guarantees, run the script manually on a copy of your data first.

Review Dimensions

Purpose & Capability
okName/description (CSV profiling and cleanup planning) align with the included artefacts: a small Python helper script and a checklist resource. Declared runtime requirement is only python3, which is proportionate to the stated purpose.
Instruction Scope
okSKILL.md restricts behavior to profiling, schema-normalization guidance, and producing plans/preview artifacts. It references only the local script and resource file. The bundled script only reads the explicitly provided CSV path and writes a JSON profile output; SKILL.md emphasizes preview-first and avoiding destructive actions unless the user asks.
Install Mechanism
okNo install spec is present (instruction-only skill with a local script). This is low-risk: nothing is downloaded or executed from remote hosts and the helper script is local and auditable.
Credentials
okThe skill declares no environment variables, credentials, or config paths. The Python script does not access environment secrets or external services. Requested inputs (CSV path, target schema, etc.) match the functionality.
Persistence & Privilege
okalways is false and the skill does not request persistent/privileged agent presence or modify other skills. Autonomous invocation remains platform-default and is not a specific red flag here.