godkiller-mcp
An MCP server that enforces hard quality gates on AI coding agents, including forced web search, planning, and empirical test execution, to prevent common failure modes and ensure production-grade code.
README
GODKILLER MCP
Phase kernel for Google Antigravity.
Plan. Gate. Prove on disk. Then claim done — or get blocked.
pip install godkiller-mcp
Contact: Facebook · Instagram @Kayvin.th
Mission
Antigravity skips phases. GODKILLER does not allow that.
| Failure mode | Kernel response |
|---|---|
No /plan |
plan / phase gates |
| Multi-file rush | /ultradeep one-file think → plan → edit |
| Fake “done” | verify_bundle + claim_done block |
| Session amnesia | marathon + memory graph |
goal → mode → evidence/plan → gated edit → disk verify → claim_done
Arena — Bare vs GODKILLER
| Control | Value |
|---|---|
| Runtime | Gemini 3.6 Flash (HIGH) — identical both arms |
| Bare | 3_WITHOUT_MCP |
| GODKILLER | 2_WITH_MCP |
| Oracle | sealed hidden_oracle/ pytest |
Single variable: MCP off / on.
Artifacts: 11_dimension_scorecard.md · 5_dimension_audit_log.json
Gauntlet
| Gate | Load |
|---|---|
| Tier 1 | ×50 |
| Tier 2 | ×150 |
| Tier 3 | ×300 |
| Nightmare | enterprise deadlock / ledger stress |
| TAU-style SOTA | state-drift / rate-limit / locks |
| Blind oracle | disk green required |
Scorecard (11)
| # | Dimension | Bare | GODKILLER | Winner |
|---|---|---|---|---|
| 1 | Pass rate | 516 / 516 | 516 / 516 | Tie |
| 2 | Speed | 0.36–0.37s | 0.31–0.32s (−16.2%) | GODKILLER |
| 3 | Tokens | ~35k–46k | ~50k–60k | Bare |
| 4 | Diff mass | +59 −52 | +73 −54 | GODKILLER |
| 5 | AST nodes | 2,840 | 3,120 (+9.8%) | GODKILLER |
| 6 | Anti-fake-claim | summary without green | live pytest gate | GODKILLER |
| 7 | Read scope | partial skim | full-scope gate | GODKILLER |
| 8 | Council | none | Coder / Hacker / Optimizer | GODKILLER |
| 9 | Rules | none | AGENTS.md + phases | GODKILLER |
| 10 | Defense | local patch | guards + type bounds | GODKILLER |
| 11 | Memory | .txt |
marathon + graph | GODKILLER |
Pass-rate tie. Process dominance: GODKILLER. Token premium: accepted.
Kernel
| Layer | Enforcement |
|---|---|
| Phase machine | assert_phase / claim_done |
| Plan OS | 9-step validate before fix edits |
| Edit safety | blast radius + check_edit_safe |
/ultradeep |
one phase/turn · one file/edit cycle |
| Verify | disk commands only |
| Memory | task → phase → evidence → lesson |
| Surface | gk_code / gk_scan / gk_browser |
/ultradeep
queue → think (≥3 hypotheses) → plan → edit_safe (1 path) → verify → advance
Install
pip install godkiller-mcp
# pip install 'godkiller-mcp[browser]' && playwright install chromium
# pip install 'godkiller-mcp[scrape]'
{
"mcpServers": {
"godkiller": { "command": "godkiller-mcp" }
}
}
Alternate: python -m godkiller_mcp.server
Optional: GODKILLER_TOOLS_DIR
Facades (12)
| Tool | Domain |
|---|---|
gk_route |
/ask /plan /debug /ultradeep /verify |
gk_mode |
protocol · skills · file gate |
gk_task |
open · blast · edit_safe |
gk_phase |
assert · claim_done · rubric |
gk_evidence |
submit · shot · critic · journey |
gk_verify |
bundle · soak · loop · competitor |
gk_memory |
lessons · marathon · graph |
gk_code |
map · search · read_full · council |
gk_scan |
AST/CWE · Semgrep |
gk_browser |
navigate · snapshot · click · fill |
gk_handoff |
spec · feedback |
gk_meta |
plan_template · plan_validate |
Modes
| Mode | Contract |
|---|---|
/ask |
no application edits |
/plan |
9-step before build |
/debug |
repro + hypothesis before fix |
/ultradeep |
marathon · max tools · per-file gate |
/verify |
proof → claim_done |
Security
- Secret values never returned over MCP (names only)
- Prefer
shell=Falseverify/soak - Network scrape/browser are explicit invocations
License
MIT © 2026 GODKILLER Team — LICENSE
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.