Propose a patch

A fix the maintainer can act on, with the evidence attached. It opens under your GitHub account in your browser; KillerSkills holds no token and never modifies the package. The fix is verified the only way we verify anything: the next crawl re-runs the new version in the honeypot. 73 · Caution

Evidence

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

Diff · removes 6 caught lines

Download .patch

diff --git a/scripts/update_docs.py b/scripts/update_docs.py
--- a/scripts/update_docs.py
+++ b/scripts/update_docs.py
@@ -96,1 +96,0 @@
-"""Update CLAUDE.md Recent Activity section.
@@ -158,1 +157,0 @@
-print(f"{YELLOW}[DRY RUN]{RESET} Would add to CLAUDE.md Recent Releases:")
@@ -168,1 +166,0 @@
-"""Update the 'Last Updated' timestamp in CLAUDE.md.
@@ -189,1 +186,0 @@
-f"{YELLOW}[DRY RUN]{RESET} Would update CLAUDE.md Last Updated to {current_date}"
@@ -224,1 +220,0 @@
-# Update CLAUDE.md only for minor/major releases or significant patches
diff --git a/scripts/reorganize-docs.sh b/scripts/reorganize-docs.sh
--- a/scripts/reorganize-docs.sh
+++ b/scripts/reorganize-docs.sh
@@ -181,1 +181,0 @@
-echo "5. Update README.md and CLAUDE.md"

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 4.1.14. 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-vector-search: Instructs the agent to modify its persistent memory or instruction files (KillerSkills verdict caution 73)

## 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 73/100**. Summary: no network, touches nothing outside the project, writes no memory.

- See the findings below.

Full trace and findings: https://killerskills.ai/skills/pypi/mcp-vector-search/4.1.14

## Findings with a location
- `scripts/update_docs.py:224` · Instructs the agent to modify its persistent memory or instruction files — scripts/update_docs.py:224 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `scripts/reorganize-docs.sh:181` · Instructs the agent to modify its persistent memory or instruction files — scripts/reorganize-docs.sh:181 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `scripts/update_docs.py:96` · Instructs the agent to modify its persistent memory or instruction files — scripts/update_docs.py:96 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `scripts/update_docs.py:189` · Instructs the agent to modify its persistent memory or instruction files — scripts/update_docs.py:189 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `scripts/update_docs.py:168` · Instructs the agent to modify its persistent memory or instruction files — scripts/update_docs.py:168 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `scripts/update_docs.py:158` · Instructs the agent to modify its persistent memory or instruction files — scripts/update_docs.py:158 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file

## Proposed change
The attached patch removes the 6 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._

Open the issue on GitHub (bobmatnyc/mcp-vector-search)Fork to open a PR

# or from a terminal, under your own account:
gh repo fork bobmatnyc/mcp-vector-search --clone && cd mcp-vector-search
curl -s https://killerskills.ai/skills/pypi/mcp-vector-search/4.1.14/patch.diff | git apply --unidiff-zero
git checkout -b killerskills/4.1.14 && git commit -am "mcp-vector-search: Instructs the agent to modify its persistent memory or instruction files"
gh pr create --fill --body-file <(curl -s https://killerskills.ai/skills/pypi/mcp-vector-search/4.1.14/patch.md)

We count proposals opened, not merged: a merge shows up as a new version that the crawl re-runs. Disputes: the maintainer can contest the verdict from its page; a person reviews it and the outcome is published.