Propose a patch

Evidence

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

diff --git a/README.md b/README.md
--- a/README.md
+++ b/README.md
@@ -34,1 +34,0 @@
-<summary><b>Table of contents</b></summary> - [Safety and privacy](#safety-and-privacy) - [How it works, in one picture](#how-it-works-in-one-picture) - [Words 
diff --git a/tests/test_auth.py b/tests/test_auth.py
--- a/tests/test_auth.py
+++ b/tests/test_auth.py
@@ -74,1 +74,0 @@
-("http://claude.ai/cb", False), ("https://evil.com\\@claude.ai/cb", False), ("https://[email protected]/cb", False), ("https://user:[email protected]/cb", False)])
diff --git a/docs/faq.md b/docs/faq.md
--- a/docs/faq.md
+++ b/docs/faq.md
@@ -9,1 +9,0 @@
-<summary><b>Is it free?</b></summary> Yes. The software is free and open source (Apache-2.0). You run it on a server you already have. Your AI app may need its 
@@ -30,1 +29,0 @@
-<summary><b>Who is it for?</b></summary> People who run apps on their own server with Coolify and keep their domains on Cloudflare, and want an AI to help from 
diff --git a/tests/test_round3.py b/tests/test_round3.py
--- a/tests/test_round3.py
+++ b/tests/test_round3.py
@@ -298,1 +298,0 @@
-task = asyncio.create_task(CALL("coolify_env_set", {"id": APP, "key": "AUTH_SERVER_URL", "value": "https://user:[email protected]/x"}))
@@ -378,1 +377,0 @@
-assert terminal.visible("ok\n‮​x\u0007") == "ok\\n\\u202e\\u200bx\\u0007"
diff --git a/.github/PULL_REQUEST_TEMPLATE.md b/.github/PULL_REQUEST_TEMPLATE.md
--- a/.github/PULL_REQUEST_TEMPLATE.md
+++ b/.github/PULL_REQUEST_TEMPLATE.md
@@ -19,1 +19,0 @@
-Does this send data somewhere new, store something new, or let the AI do more without asking?

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 0.1.0. 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: coolify-cloudflare-mcp: Collapsed <details> block carrying instructions (KillerSkills verdict caution 72)

## 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 72/100**. Summary: no network, touches nothing outside the project, writes no memory.

- See the findings below.

Full trace and findings: https://killerskills.ai/skills/glama/coolify-cloudflare-mcp/0.1.0

## Findings with a location
- `README.md:34` · Collapsed <details> block carrying instructions — README.md:34 · Collapsed <details> block carrying instructions
- `tests/test_auth.py:74` · URL with embedded credentials — tests/test_auth.py:74 · URL with embedded credentials — may be phishing or credential leaking
- `docs/faq.md:30` · Collapsed <details> block carrying instructions — docs/faq.md:30 · Collapsed <details> block carrying instructions
- `docs/faq.md:9` · Collapsed <details> block carrying instructions — docs/faq.md:9 · Collapsed <details> block carrying instructions
- `tests/test_round3.py:378` · Hidden zero-width characters detected — tests/test_round3.py:378 · Hidden zero-width characters detected — may conceal content
- `tests/test_round3.py:298` · URL with embedded credentials — tests/test_round3.py:298 · URL with embedded credentials — may be phishing or credential leaking
- `.github/PULL_REQUEST_TEMPLATE.md:19` · HTML comment carrying imperative instructions to the agent — .github/PULL_REQUEST_TEMPLATE.md:19 · HTML comment carrying imperative instructions to the agent · informational: not an instruction file

## Proposed change
The attached patch removes the 7 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._