Propose a patch

Evidence

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

diff --git a/CHANGELOG.md b/CHANGELOG.md
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -74,1 +74,0 @@
-- mailgun-mcp: Update CLAUDE.md + rationale for round 1 fixture deletion
diff --git a/RULES.md b/RULES.md
--- a/RULES.md
+++ b/RULES.md
@@ -368,1 +368,0 @@
-- AI assistants should update CLAUDE.md and RULES.md weekly or after pre-commit failures
@@ -378,1 +377,0 @@
-- Update the "AI Code Generation Best Practices" checklist in CLAUDE.md

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 0.5.4. 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: mailgun-mcp: Instructs the agent to modify its persistent memory or instruction files (KillerSkills verdict caution 78)

## What the honeypot observed
KillerSkills ran **[email protected]** in an isolated sandbox (decoy credentials, recording proxy, syscall trace; engine 1.7) 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/mailgun-mcp/0.5.4

## Findings with a location
- `CHANGELOG.md:74` · Instructs the agent to modify its persistent memory or instruction files — CHANGELOG.md:74 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `RULES.md:378` · Instructs the agent to modify its persistent memory or instruction files — RULES.md:378 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `RULES.md:368` · Instructs the agent to modify its persistent memory or instruction files — RULES.md:368 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file

## Proposed change
The attached patch removes the 3 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._