MNCS Forge MCP
An experimental MCP server for MNCS-native development and evidence control, enabling Codex to manage project authority, candidate lineage, evidence gaps, and evaluator-mode boundaries.
README
MNCS Forge MCP
MNCS Forge is an experimental, non-normative Model Context Protocol server and CLI that gives Codex an MNCS-native development and evidence control plane. It makes project authority, candidate lineage, epochs, declared checks, evidence gaps, selection, freeze, and evaluator-mode boundaries explicit.
Forge is not required for MNCS conformance, an accredited certification system, a source of independence or protected custody, or another universal Code Property Graph system. It does not replace Joern. Joern, compilers, analyzers, benchmarks, mutation systems, sanitizers, and runtime harnesses remain replaceable evidence providers.
Version 0.1.0a1 is a reference experiment. Local results do not promote MNCS, MNCDS, RFCs, or
case studies. REVIEW_REQUIRED is a workflow disposition, not an MNCS result. Status aggregation
is FAIL > UNKNOWN > PASS; absent or unsupported evidence never silently becomes PASS.
Architecture
flowchart LR
Codex[Codex / MCP client] -->|MCP stdio| Forge[MNCS Forge control plane]
Human[Human CLI] --> Forge
Forge -->|declared argv only| Providers[Provider Protocol 0.1 providers]
Forge -->|declared workflows| Harnesses[compilers, tests, benchmarks, analyzers]
Providers --> Records[append-only local records]
Harnesses --> Records
Forge --> Records
Records -->|public commands and schemas| Validators[MNCS and MNCDS offline validators]
Validators --> Separate[separate MNCS / MNCDS results]
MCP is the interactive Codex interface. Provider Protocol 0.1 is the deterministic analyzer interface. MNCS and MNCDS validators remain offline result authorities. Forge never copies their conformance logic into its control plane.
Quick start
python3 -m venv .venv
. .venv/bin/activate
python -m pip install -e '.[dev]'
mncs-forge --config examples/minimal/mncs-forge.toml config validate
mncs-forge --config examples/minimal/mncs-forge.toml inspect
For an MNCS checkout containing the committed EdgeStream integration configuration:
./scripts/install-codex-mcp.sh \
/absolute/path/to/machine-native-complexity-standard/mncs-forge.toml
./scripts/verify-codex-mcp.sh \
/absolute/path/to/machine-native-complexity-standard/mncs-forge.toml
Start a new Codex session after registration; an already-running process is not assumed to reload
MCP servers dynamically. Uninstall with ./scripts/uninstall-codex-mcp.sh.
Interfaces
The mncs-forge CLI provides doctor, inspect, status, blockers, epoch begin,
candidate register/compare/select/reject, check development, explain, freeze, evaluate,
reconcile, bundle, ledger verify, and config validate.
The mncs-forge-mcp stdio server exposes project inspection, separate claim status and blockers,
epoch/candidate control, declared checks, compact failure explanation, policy-bound comparison and
disposition, freeze/final evaluation, reconciliation, and bundle orchestration. Read-only
resources summarize authority, active state, evidence, blockers, and usage. Prompts guide the
controlled workflow but cannot bypass tool authority checks.
Documentation
- Architecture and trust boundaries
- Configuration
- Security model and residual risks
- CLI and MCP surface
- Codex installation
- EdgeStream integration
- Provider Protocol integration
- Evidence and identity model
- Compatibility
Licensed under Apache-2.0.
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.