unifi-mcp-shared@0.6.43
VERIFIED· install onlyPublisher claims: "Shared MCP server patterns: permissions, confirmation, lazy loading, config" Observed in sandbox: no network, touches nothing outside the project, writes no memory.
Coverage
Bundled scripts ran with no arguments; no agent task scenario for this bundle.- Tested
- 8 of 26 bundled scripts run with no arguments (the rest were not run: per-bundle cap), with decoy credentials in the environment · every file read, file written and host contacted by what ran
- Not tested
- use of the skill by an agent
What this means for the label: Verified here covers what ran. The server's tools were not exercised with a real credential, so nothing about their behavior is asserted; the credential summary below says what to do with that.
Before you give it your token
- It will ask for
UNIFI_NETWORK_PASSWORDUNIFI_PREVIEW_TOKEN_TTL_SECONDSConfiguration, not credentials: CONFIG_PATH, OTEL_SDK_DISABLED, SystemRoot, UNIFI_MCP_PROTOCOL_REVISION, UNIFI_MCP_PROTOCOL_VERSION, UNIFI_TOOL_REGISTRATION_MODE- Where it talks to
- no vendor host named in the code
- What we saw when it ran
- no credential files read · no network
- What we could not test
- use of the skill by an agent
Connect it with a scoped credential, under the guard. It asks for UNIFI_NETWORK_PASSWORD, UNIFI_PREVIEW_TOKEN_TTL_SECONDS and talks to its vendor. Give the credential to the guard, not the server: the value never enters the server's environment and only travels to the host you bind. What the tools do with a real credential was not observed.
Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.
What changed since 0.6.40
vet-skills diff unifi-mcp-shared 0.6.40 0.6.43In the sandbox, this version reached for nothing 0.6.40 did not. Compared from the two recorded runs, not from source.
The guard can hold an update like this until someone approves it (runtime.on_change: hold). A change that only shows with a real credential, or on a path the run did not reach, is not in a trace and so not in this comparison.
Is [email protected] safe to install?
KillerSkills ran it in the KillerSkills honeypot sandbox (2 scenarios; bundled scripts ran with no arguments; no agent task scenario for this bundle) on 2026-11-20 and scored it 81/100 (Verified). In what ran, nothing we could detect went wrong: no undeclared network egress, no credential reads, no persistent instruction writes. What it does with a real credential was not observed; the static findings and the publisher carry the rest of the score. Summary: no network, touches nothing outside the project, writes no memory.
Behavior trace
Honeypot run · 1 s · 2 scenarios · 0 deviationsStatic findings
This is not project ``.env`` discovery: the path is set explicitly by theFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.its ``.env`` would let that project choose credential destinations, policy,Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.f"- Codex or OpenClaw: run the {plugin_name}-setup skill. It writes ~/.config/unifi-mcp/{product}.env, "Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose."""A ``.env`` in the working directory must not be able to point the server at a file via ``_FILE``.Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.(tmp_path / ".env").write_text(dotenv, encoding="utf-8")Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.assert not marker.exists(), "a .env-supplied _COMMAND was executed"Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.server never discovers a project ``.env``, so the names it sets must notFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose."""The variable name, tagged when a .env supplied it rather than the process environment."""Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.f"- Codex or OpenClaw: run the {plugin_name}-setup skill. It writes ~/.config/unifi-mcp/{product}.env, "Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose."""A ``.env`` in the working directory must not be able to make the server run a program.Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.loads project ``.env`` files, so this is a second line rather than the first.Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.return f"{var} (from a .env file)"Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose."%s was supplied by a .env file, not by the environment the server was started with; "Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.f"subprocess.Popen([sys.executable, '-c', {descendant!r}])\n"Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.proc = subprocess.Popen( # noqa: S603 -- argv from the operator's own environment, never a shellFix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.result = subprocess.run([sys.executable, str(probe)], capture_output=True, timeout=30, check=False)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.# Simulate: process started without the variable, then a .env in the cwd set it.Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.proc = subprocess.Popen(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.assert ".env" in caplog.textFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose."""Both names are logged, the .env-supplied one tagged, and neither value."""Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.assert ".env" in caplog.textFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.(tmp_path / ".env").write_text("UNIFI_NETWORK_HOST=project.invalid\n")Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out."[credentials] %s points at %s with mode %04o; consider chmod 600 so other users cannot read it.",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.return (logging.WARNING, "with mode %04o; consider chmod 600 so other users cannot read it.", mode)Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.monkeypatch.setenv("UNIFI_NETWORK_PASSWORD", "plain") # left over in a copied .envFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose."""A .env arriving after the snapshot must not reach the helper's environment."""Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.return subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.proc = subprocess.Popen(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.result = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.return (logging.ERROR, "with mode %04o, which other users can write; ignoring it. Run chmod 600.", mode)Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.``.env`` may have just rewritten -- which would defeat the name filter.Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.# Only the indirection is gated; a .env may still carry the value itself.Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.assert any(level == logging.WARNING and "chmod 600" in message for level, message, _ in report.messages)Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.proc = subprocess.Popen(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.proc = subprocess.Popen(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.(bundled_config / ".env").write_text("UNIFI_HOST=attacker.invalid\n")Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.proc = subprocess.Popen(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.so that a failed HTTP transport does not kill the stdio connection.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.proc = subprocess.Popen(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.Cost to the agent
Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.
Known vulnerabilities
Not checked against OSV yet; the weekly check has not reached this version.