Skill Hub

ReviewAudited by ClawScan on May 10, 2026.

Overview

The skill mostly matches its stated purpose, but its vetting command can be steered into scanning files outside the skills folder if given a malformed skill name.

Use this skill only in a trusted workspace, do not vet slugs containing slashes or '..', review results before installing any skill, and consider pinning/verifying the ClawHub CLI and catalog source. Treat PASS/Trusted labels as helpful signals, not proof that a skill is safe.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

A malformed or malicious skill name could cause the vetter to scan unintended local directories and expose file paths or matched snippets to the agent output.

Why it was flagged

The user- or catalog-provided slug is joined into a local path without rejecting absolute paths, separators, or '..', then passed to the recursive scanner.

Skill content
parser.add_argument("--slug", "-s", help="Vet a single skill by name")
...
local = WORKSPACE / "skills" / slug
if local.exists():
    return local, False
...
findings = scan_skill_dir(skill_path)
Recommendation

Validate slugs with a strict allowlist, resolve the path, reject '..' and path separators, and verify the final path remains inside the intended skills directory before scanning.

What this means

Registry operations may execute whatever version of the ClawHub CLI is current when the command runs.

Why it was flagged

The skill invokes a mutable latest-version external CLI at runtime for registry operations, which is expected for this integration but depends on external package provenance.

Skill content
["npx", "clawhub@latest", "list"]
Recommendation

Prefer a pinned ClawHub CLI version or verify the CLI source before use, especially before installing or inspecting unknown skills.

What this means

GitHub API checks may use your existing gh login and its permissions.

Why it was flagged

The quick-check feature uses the GitHub CLI and may run under the user's logged-in GitHub identity, although no credential theft or logging is shown.

Skill content
output = run_gh([
    "api", f"repos/{REPO}/commits?per_page=1",
...
print("  Could not reach GitHub API. Check `gh auth status`.")
Recommendation

Confirm which GitHub account gh is authenticated as, and use a least-privileged token/account if possible.

What this means

If the remote catalog contains misleading text or maliciously crafted entries, that content can persist locally and influence later recommendations.

Why it was flagged

The skill stores mutable remote catalog content locally and later uses it for search results, recommendations, and vetting targets.

Skill content
RAW_URL = "https://raw.githubusercontent.com/VoltAgent/awesome-openclaw-skills/master/README.md"
...
"skills": new_skills,
...
json.dump(catalog, f, indent=2)
Recommendation

Treat catalog entries as untrusted data, refresh from trusted sources, and avoid batch-vetting or installing catalog entries without review.