Skillv2.0.0

ClawScan security

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

Scanner verdict

BenignMar 17, 2026, 6:59 PM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's files and runtime instructions align with an .env/template/validation tool — it reads and writes local env files and stores logs in a per-user data directory, but does not demand external credentials or perform network exfiltration.
Guidance: This skill appears to be what it says: a local .env/template/validation tool. Before using it, review the scripts (they are included), and be aware: 1) invoking validate/encrypt/decrypt reads your .env files (which often contain secrets); 2) encrypt uses Base64 (not real encryption) and will expose encoded secrets on stdout — do not treat that as secure storage; 3) the tool creates a per-user data dir (~/.local/share/env-config by default) and logs history/data, so sensitive values may be written to disk; 4) if you want to limit persistence, set ENV_CONFIG_DIR to a controlled location or inspect/clean the data directory after use. If you are comfortable with those behaviors, the skill is coherent and usable; if not, run it in an isolated container or inspect/modify the scripts before running.

Review Dimensions

Purpose & Capability: okName, description, SKILL.md and the two scripts implement .env template generation (Node/Python/Go/Docker/etc.), validation, compare, base64 encode/decode, gitignore snippet, and a small local data/logger — all coherent with an environment-variable config manager.
Instruction Scope: noteSKILL.md and scripts focus on .env tasks. The scripts explicitly read .env and .env.example, can base64-encode/decode their contents, and provide comparison/validation. This is within the stated scope, but be aware the tool will read any .env file you point it at (which may contain secrets) and will print encoded/decoded content to stdout.
Install Mechanism: okInstruction-only skill (no install spec). Provided scripts are local and do not download or execute remote artifacts. No high-risk install mechanism observed.
Credentials: noteThe skill does not request credentials or special env vars, but the scripts honor ENV_CONFIG_DIR/XDG_DATA_HOME/HOME and will create/read files under the user's data directory. Templates include placeholder names for services (AWS, OPENAI, STRIPE) as expected, but the tool will read .env files that may contain real secrets — so access to those files is functionally required but sensitive.
Persistence & Privilege: notealways:false and no platform-level privileges. The scripts create a per-user data directory (default: ${XDG_DATA_HOME:-$HOME/.local/share}/env-config), write history.log and data.log, and may persist entries added via commands. This is reasonable for a CLI utility but means user data and command history are stored on disk.