Propose a patch

Evidence

The sandbox observed nothing beyond the static findings; the proposal cites those.

diff --git a/src/mcp_relay/auth.py b/src/mcp_relay/auth.py
--- a/src/mcp_relay/auth.py
+++ b/src/mcp_relay/auth.py
@@ -1,1 +1,0 @@
-"""Authentication helpers for untrusted relay credentials."""
diff --git a/src/mcp_relay/config.py b/src/mcp_relay/config.py
--- a/src/mcp_relay/config.py
+++ b/src/mcp_relay/config.py
@@ -1906,1 +1906,0 @@
-"""Mask relay tokens and credential-like keys in config show/get output.
diff --git a/tests/test_dotenv_tolerant.py b/tests/test_dotenv_tolerant.py
--- a/tests/test_dotenv_tolerant.py
+++ b/tests/test_dotenv_tolerant.py
@@ -3,1 +3,0 @@
-Any entry except the relay credentials becomes a process environment
diff --git a/tests/test_provider_tools.py b/tests/test_provider_tools.py
--- a/tests/test_provider_tools.py
+++ b/tests/test_provider_tools.py
@@ -590,1 +590,0 @@
-"resource": {"uri": "https://user:[email protected]/x"},
@@ -598,1 +597,0 @@
-credential_uri = "https://user:[email protected]/resource?token=top-secret"

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 0.1.1. If the behavior is intended, declare it (host under permissions.network, a line in the README) instead of removing it; the next crawl scores declarations.

Issue / PR body

Title: mcp-relay: Credential pass-through configuration (KillerSkills verdict caution 68)

## What the honeypot observed
KillerSkills ran **[email protected]** in an isolated sandbox (decoy credentials, recording proxy, syscall trace; engine 1.8.1) and scored it **CAUTION 68/100**. Summary: calls astral.sh only, touches nothing outside the project, writes no memory.

- See the findings below.

Full trace and findings: https://killerskills.ai/skills/pypi/mcp-relay/0.1.1

## Findings with a location
- `src/mcp_relay/auth.py:1` · Credential pass-through configuration — src/mcp_relay/auth.py:1 · Credential pass-through configuration — agent relays sensitive credentials
- `src/mcp_relay/config.py:1906` · Credential pass-through configuration — src/mcp_relay/config.py:1906 · Credential pass-through configuration — agent relays sensitive credentials
- `tests/test_dotenv_tolerant.py:3` · Credential pass-through configuration — tests/test_dotenv_tolerant.py:3 · Credential pass-through configuration — agent relays sensitive credentials
- `tests/test_provider_tools.py:590` · URL with embedded credentials — tests/test_provider_tools.py:590 · URL with embedded credentials — may be phishing or credential leaking
- `tests/test_provider_tools.py:598` · URL with embedded credentials — tests/test_provider_tools.py:598 · URL with embedded credentials — may be phishing or credential leaking

## Proposed change
The attached patch removes the 5 lines above. If the behavior is intended, the fix is to declare it instead: list the host under `permissions.network` (or the MCP manifest) and say in the README why the package needs it; the next crawl will re-score it with the declaration taken into account.

## How this will be verified
Nothing is "verified safe" by assertion. When a new version is published, KillerSkills re-runs it in the honeypot and the verdict page updates; you can trigger it early from the verdict page ("Re-scan after I publish a fix"). Reply here or at [email protected] if you believe the verdict is wrong; disputes are reviewed by a person and the outcome is published on the verdict page.

_Proposed through killerskills.ai/threats by a user of the package, under their own GitHub account. The evidence is public; the package was never modified by KillerSkills._