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
@@ -136,1 +136,0 @@
-<summary><b>MCP Server — Claude Code / Claude Desktop / Cursor</b></summary> From a clone (reads `.env` in the repository): ```bash claude mcp add gs-robot -- u
@@ -162,1 +161,0 @@
-<summary><b>HTTP Server + H5 mobile app</b></summary> ```bash uv run gs-robot-server # http://0.0.0.0:8000 (GS_SERVER_HOST / GS_SERVER_PORT) ``` - **H5 app** — 
@@ -445,1 +443,0 @@
-To add a V3 tool, define the input model and handler in `src/gs_openapi/tools/v3_tools.py` with `@tool(...)`; MCP, REST and the Agent pick it up automatically. Add a test and a row to the tool tables.
diff --git a/src/gs_openapi/server/static/assets/cytoscape.esm-C5JX3QLV.js b/src/gs_openapi/server/static/assets/cytoscape.esm-C5JX3QLV.js
--- a/src/gs_openapi/server/static/assets/cytoscape.esm-C5JX3QLV.js
+++ b/src/gs_openapi/server/static/assets/cytoscape.esm-C5JX3QLV.js
@@ -14,1 +14,0 @@
-*/var Yv=function(e,t){this.recycle(e,t)};function ca(){return!1}function ja(){return!0}Yv.prototype={instanceString:function(){return"event"},recycle:function(e,t){if(this.isImmediatePropagationStopp
diff --git a/src/gs_openapi/server/static/assets/katex-C5jXJg4s.js b/src/gs_openapi/server/static/assets/katex-C5jXJg4s.js
--- a/src/gs_openapi/server/static/assets/katex-C5jXJg4s.js
+++ b/src/gs_openapi/server/static/assets/katex-C5jXJg4s.js
@@ -251,1 +251,0 @@
--470,-1265c-4.7,-6,-9.7,-11.7,-15,-17c-0.7,-0.7,-6.7,-1,-18,-1z`;default:throw new Error("Unknown stretchy delimiter.")}};function ga(r){return"toText"in r}class ae{constructor(e){this.children=void 0

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 0.4.1. 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-gs-robot: Collapsed <details> block carrying instructions (KillerSkills verdict caution 74)

## 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 74/100**. 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-gs-robot/0.4.1

## Findings with a location
- `README.md:162` · Collapsed <details> block carrying instructions — README.md:162 · Collapsed <details> block carrying instructions · informational: not an instruction file · informational: not an instruction file
- `README.md:136` · Collapsed <details> block carrying instructions — README.md:136 · Collapsed <details> block carrying instructions · informational: not an instruction file · informational: not an instruction file
- `README.md:445` · Instructs the agent to modify its persistent memory or instruction files — README.md:445 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file · informational: not an instruction file · informational: not an instr
- `src/gs_openapi/server/static/assets/cytoscape.esm-C5JX3QLV.js:14` · Hidden zero-width characters detected — src/gs_openapi/server/static/assets/cytoscape.esm-C5JX3QLV.js:14 · Hidden zero-width characters detected — may conceal content
- `src/gs_openapi/server/static/assets/katex-C5jXJg4s.js:251` · Hidden zero-width characters detected — src/gs_openapi/server/static/assets/katex-C5jXJg4s.js:251 · Hidden zero-width characters detected — may conceal content

## 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._