📡 MCP Radar

CogniSwitch/KL4A

Turn SOP, policy and regulatory documents into human-verified Open Knowledge Format (OKF) bundles — every claim carries its source span. Desktop app + CLI + MCP server.

7 stars
0 forks
Rust
momentum ▲ 17.5
created 2026-08-25
on radar since 2026-08-30
agent-skillsagentsclaude-codecompliancehuman-in-the-loopknowledge-graphsmcpmcp-serverneuro-symbolic-aiokfokf-bundleokf-knowledge-bundleokf-v0-2open-knowledge-formatprovenanceragsoptauri
View on GitHub ↗Homepage ↗

About KL4A

Create OKF-compliant knowledge bundles from your SOP docs, and enable your agents to use them. Every claim carries the exact source span it came from, and nothing is marked verified until a human approves it.

Feed it a PDF, a DOCX, or a plain-text SOP — real, unstructured prose, not a website's DOM or a codebase's AST. It then:

- Normalizes the document into clean Markdown. - Extracts obligation-shaped claims (must, shall, should record, ...) and attaches each one to the exact byte range of the source sentence it came from. - Records every approve, reject, and edit as its own event, with a reviewer, a rationale, and a before/after diff, directly in the bundle. - Populates the claim's OKF v0.2 trust fields (provenance, verified, lifecyclestatus) on approval with who verified it and when, rather than leaving them blank for someone downstream to fill in by hand.

The result is a plain-file artifact. Markdown with YAML frontmatter: readable in a text editor, diffable with git, and queryable by an agent over MCP or the CLI — with no database, and no server standing between them and the source of truth.

A native desktop app (Windows/macOS/Linux, built on Tauri) — one application to install, not a server to stand up or a runtime to provision.

Your SOPs, policies, and regulations live in prose written for a human reader, not an agent. Pasting that prose straight into a prompt means the agent is trusting a wall of text with no accountability trail — if it gets a clause wrong, nothing tells you which sentence it misread or who signed off on it.

From the project README.

Maintaining this server?

Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:

[![On MCP Radar](https://mcp.liqiwa.com/badge.svg)](https://mcp.liqiwa.com/s/CogniSwitch--KL4A.html)

Related MCP servers

menot-you/n-memory

Memory your agent can trust — every answer carries its source; when it doesn't know, it says so. Hermetic, local, MCP.

⭐ 5Rust
omaekumiko2-create/kru

Local-first MCP password and credential manager for AI agents. Use passwords, API keys, SSH identities, and TOTP without exposing hidden plaintext to the model.

⭐ 88Rust
montezuma-p/harken

Your agent cannot listen to audio. harken fixes that — locally. MCP server + CLI, one 13 MB binary, fully offline: no Python, no ffmpeg, no API key.

⭐ 12Rust
surendranb/writing-skills

Procedural writing-style skills for AI agents: official frameworks (plain language, business writing, GOV.UK, AP style, STE-100, developer docs) + character voices. Agent Plugins 1.0.0, works in openc

⭐ 19Python
MagePsycho/awesome-magento-ai

Curated list of AI tools for Magento 2, Adobe Commerce & Mage-OS — agent skills, MCP servers, modules, CLIs, starter kits

⭐ 10
KernelLord/pickysteve

Skill router and context picker for Claude Code, Cursor, and Aider. Auto-picks the right skill for every prompt, with prompt-injection filtering for MCP.

⭐ 6Python

📬 Get the weekly radar in your inbox

The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.