Propose a patch

Evidence

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

diff --git a/src/infrahub_mcp/auth.py b/src/infrahub_mcp/auth.py
--- a/src/infrahub_mcp/auth.py
+++ b/src/infrahub_mcp/auth.py
@@ -80,1 +80,0 @@
-"""Store the passthrough token for the current async task.
@@ -96,1 +95,0 @@
-"""Read the passthrough token for the current async task."""
diff --git a/src/infrahub_mcp/middleware.py b/src/infrahub_mcp/middleware.py
--- a/src/infrahub_mcp/middleware.py
+++ b/src/infrahub_mcp/middleware.py
@@ -230,1 +230,0 @@
-Rejects tool calls and resource reads when no passthrough credential is
diff --git a/tests/unit/test_auth.py b/tests/unit/test_auth.py
--- a/tests/unit/test_auth.py
+++ b/tests/unit/test_auth.py
@@ -225,1 +225,0 @@
-class TestPassthroughTokenContextVar:
diff --git a/src/infrahub_mcp/utils.py b/src/infrahub_mcp/utils.py
--- a/src/infrahub_mcp/utils.py
+++ b/src/infrahub_mcp/utils.py
@@ -131,1 +131,0 @@
-passthrough token or Basic credentials), when there is one.

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 1.3.0. 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: infrahub-mcp: Credential pass-through configuration (KillerSkills verdict caution 78)

## 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 78/100**. Summary: no network, touches nothing outside the project, writes no memory.

- See the findings below.

Full trace and findings: https://killerskills.ai/skills/pypi/infrahub-mcp/1.3.0

## Findings with a location
- `src/infrahub_mcp/auth.py:80` · Credential pass-through configuration — src/infrahub_mcp/auth.py:80 · Credential pass-through configuration — agent relays sensitive credentials
- `src/infrahub_mcp/auth.py:96` · Credential pass-through configuration — src/infrahub_mcp/auth.py:96 · Credential pass-through configuration — agent relays sensitive credentials
- `src/infrahub_mcp/middleware.py:230` · Credential pass-through configuration — src/infrahub_mcp/middleware.py:230 · Credential pass-through configuration — agent relays sensitive credentials
- `tests/unit/test_auth.py:225` · Credential pass-through configuration — tests/unit/test_auth.py:225 · Credential pass-through configuration — agent relays sensitive credentials
- `src/infrahub_mcp/utils.py:131` · Credential pass-through configuration — src/infrahub_mcp/utils.py:131 · Credential pass-through configuration — agent relays sensitive credentials

## 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._