Security audit
Security checks for vulnerabilities and agentic risk
This skill is purpose-built for publishing to UseClaw, but it needs Review because it installs an unchecked remote CLI and handles a persistent publishing token with limited safeguards.
Install only if you trust UseClaw's remote CLI delivery path. Prefer a pinned or signed CLI release before use, treat the UseClaw token as secret, avoid putting real tokens in chat or shell history, review content before publishing, and rotate or revoke the token if it may have appeared in logs.
skill.json:8Unverified Remote CLI Payload Downloaded and Installed as an Executable
SKILL.md:39UseClaw Authentication Token Passed Through a Command-Line Argument
Referencing a concrete credentials file path is not inherently malicious, but in this context it identifies where reusable secrets are stored and normalizes persistent local credential storage without any security caveats. If an agent or user is induced to inspect, copy, or mishandle that file, the stored token could be exfiltrated and reused to publish as the victim.
Credentials are stored at:
~/.config/useclaw/credentials.json
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
"cli": {
"name": "useclaw",
"download": "curl -sL https://useclaw.net/cli/useclaw-cli.sh -o ~/.local/bin/useclaw && chmod +x ~/.local/bin/useclaw",
"credentials_path": "~/.config/useclaw/credentials.json"
},
"commands": [
{
The skill directs the operator to input a personal access token on the command line and notes where credentials are stored, but it does not warn that the token is sensitive, persistent, and should not be exposed in logs, shell history, or transcripts. In an agent setting, this increases the chance that secrets are mishandled or echoed back into observable channels, which can lead to account compromise.
The skill collects a long-lived token and stores it in a local credentials file, but the manifest does not warn the user about local persistence, file location, or the sensitivity of the credential. This can lead to accidental exposure through backups, shared machines, weak file permissions, or users providing tokens without understanding they will be stored on disk.
The skill is explicitly designed to publish user-provided content to an external platform, but the manifest provides no privacy, consent, or disclosure warning before transmission. This creates a real risk that an agent may send sensitive drafts, proprietary data, or personal information off-platform without the user fully understanding that publication is external and potentially public.
No suspicious patterns detected.