ClawAudit verdict
mcp-marketplace
MCP server discovery and installation marketplace skill that searches registries and generates configs; the skill explicitly warns never to store secrets raw and guides users through auth setup, consistent with the stated server management purpose.
⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.
Automated static analysis — not a human review. ClawAudit flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.
Findings (7)
Accesses shell history/config
references/transport-patterns.md · code · ~/.zshrc
Pipe-to-shell pattern (curl | sh) — supply chain attack vector
references/troubleshooting.md · prose · downgraded · curl -LsSf https://astral.sh/uv/install.sh | sh
Pipe to sh — executes piped content as shell commands
references/troubleshooting.md · prose · downgraded · | sh
subprocess execution — runs system commands from Python
scripts/build_config.py · prose · downgraded · subprocess.run(
Accesses system credential store
scripts/secrets_helper.py · prose · downgraded · Keychain
Python os.environ.get — reads environment variable
scripts/build_config.py · prose · downgraded · os.environ.get(
Python urllib.request — network access
scripts/health_check.py · prose · downgraded · urllib.request
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class A/D/F). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.
Permissions & capabilities
No declared permissions — minimal attack surface.
network_in Is this flag fair?
Thanks — recorded.