Propose a patch

Evidence

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

diff --git a/src/mcpbait/modules/elicitation_phish.py b/src/mcpbait/modules/elicitation_phish.py
--- a/src/mcpbait/modules/elicitation_phish.py
+++ b/src/mcpbait/modules/elicitation_phish.py
@@ -36,1 +36,0 @@
-"let an agent relay credential requests; route authentication through a channel "
diff --git a/docs/LEADERBOARD.md b/docs/LEADERBOARD.md
--- a/docs/LEADERBOARD.md
+++ b/docs/LEADERBOARD.md
@@ -112,1 +112,0 @@
-3. **Workspace Rulefile Write Protection:** Treat `.cursorrules`, `CLAUDE.md`, and `AGENTS.md` as privileged artifacts requiring explicit, user-confirmed modal approvals before modification.
diff --git a/src/mcpbait/modules/memory_poisoning.py b/src/mcpbait/modules/memory_poisoning.py
--- a/src/mcpbait/modules/memory_poisoning.py
+++ b/src/mcpbait/modules/memory_poisoning.py
@@ -53,1 +53,0 @@
-"CLAUDE.md (or .cursorrules / AGENTS.md if present). Append it verbatim, "

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 0.4.2. 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: mcpbait: Credential pass-through configuration (KillerSkills verdict caution 77)

## 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 77/100**. Summary: no network, touches nothing outside the project, writes no memory.

- See the findings below.

Full trace and findings: https://killerskills.ai/skills/pypi/mcpbait/0.4.2

## Findings with a location
- `src/mcpbait/modules/elicitation_phish.py:36` · Credential pass-through configuration — src/mcpbait/modules/elicitation_phish.py:36 · Credential pass-through configuration — agent relays sensitive credentials
- `docs/LEADERBOARD.md:112` · Instructs the agent to modify its persistent memory or instruction files — docs/LEADERBOARD.md:112 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `src/mcpbait/modules/memory_poisoning.py:53` · Instructs the agent to modify its persistent memory or instruction files — src/mcpbait/modules/memory_poisoning.py:53 · 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._