Skip to content

AI / LLM agents: For a machine-readable summary of this site, see /llms.txt or /llms-full.txt.

KL4A

Knowledge Layer For Agents — an independent open-source project.

Create OKF-compliant knowledge bundles from your SOP docs (sopkb) and your source code (codekb) — and enable your agents to use them.

→ See KL4A in Action — a real source, real commands, real output, and a real agent answering a real question, in about five minutes.

Your organization's SOPs, policies, procedures, regulations, and standards live in Word docs, PDFs, and wikis — written for humans, not for agents. KL4A reads them, proposes structured knowledge, and puts a person in front of every claim before an agent can act on it. What comes out the other side is a Knowledge Bundle:

  • Every claim traces back to the exact source text it came from.
  • Nothing an agent can act on until a person has reviewed it.
  • Plain files, git-diffable, no lock-in — built on OKF, the open format this project implements. (KL4A is not affiliated with or endorsed by the OKF project.)

It doesn't claim to mine everything in a document — nothing becomes usable knowledge until a person reviews it, which is exactly the point.

The same discipline applies to a source repository — see Code Bundles.

Ground it. Review it. Then trust it.

Built for knowledge engineers

KL4A is for the person who owns turning a folder of SOPs into something an agent can be trusted against: ingest a document set, inspect what got extracted and why, work through review until the bundle is clean, and hand it off — to an agent, an MCP server, or an enterprise import.

You review the knowledge once, so you don't have to review every answer forever.

That review happens once, at authoring time, on a bounded set of extracted claims — not at runtime, on every answer an agent gives. It's also usable without a knowledge engineer in the room:

  • SOP & policy owners see exactly what was extracted, verify it against source evidence, and approve, reject, or correct it — no CLI required.
  • Agent developers search knowledge items, resolve citations, and retrieve source-grounded context locally, via CLI, MCP, or the bundle's plain files.

What it does

  • Structure — sources become OKF-native bundles agents can query directly.
  • Ground — every claim keeps its exact source span, or is flagged when one can't be matched.
  • Review — nothing becomes accepted knowledge until a person approves, rejects, defers, edits, or comments on it.
  • Consume — agents query the bundle via CLI, agent chat, or MCP; export to Graph JSON or RDF/TTL when other tooling needs it.

One product, two options

KL4A is the product. It offers two options over one shared OKF core — pick the one that matches what you are grounding. The published kl4a binary is the entry point for CodeKB; the release also includes sopkb-cli and sopkb-mcp for the SOP workflow.

sopkb codekb
For SOPs, policies, procedures, regulations Source code
Input Word docs, PDFs, Markdown A Python or COBOL repository
Scope — Python and COBOL parsed; other languages inventoried, not parsed
Produces Claims grounded in source text Claims grounded in source spans, plus modules, symbols and relations
Uses an LLM Optionally Optionally, for hybrid mining
Start here Quickstart Getting Started

Both write OKF bundles and expose the result over a CLI and an MCP server. They differ in what goes in and what gets extracted. SOPKB has the released desktop workflow; CodeKB is available through CLI and MCP.

One CLI entry point: kl4a

Download kl4a from the latest release, put it on your PATH, then select the workflow on the command line:

kl4a --use sopkb <command>
kl4a --use codekb <command>

For backward-compatible SOPKB integrations, the same release also provides the standalone sopkb-cli and sopkb-mcp binaries.

The dispatcher isn't a fourth option. It has no behavior beyond deciding which of the two to call: same subcommands, same flags, same --help, same exit codes, the same objects when imported. Skip it and run sopkb/codekb directly if you'd rather — nothing about what they do changes either way.

→ kl4a: the unified CLI and library

Install, run, see output

No LLM, no API key, no network call

The fixture mining provider extracts obligation-shaped sentences ("must", "shall", "should record"...) with plain pattern matching, and it's the default — nothing below needs azure-llm or any credentials.

KL4A is the product; sopkb is one of its two CLIs

KL4A ships as one PyPI project, kl4a — a single pip install kl4a gets you the shared substrate plus the sopkb and codekb packages and CLIs described above, all together. The distribution once named kl4a is now sopkb: naming one option after the whole product read wrong once the shared substrate needed a real name of its own, which is kl4a today. kl4a is not a fourth option and not that old renamed distribution — besides being the substrate the two options build on, it's also the dispatcher the commands below use (kl4a --use sopkb ...) and an import namespace (kl4a.sopkb) — see kl4a: the unified CLI and library. Every kl4a --use sopkb <command> below is sopkb <command> untouched — same flags, same output, same exit code — so sopkb <command> still works exactly as shown if you'd rather type that.

pip install kl4a
kl4a --use sopkb init demo-bundle

No server. No database. No LLM key required — that's the entire dependency footprint.

From there it's scan → normalize → mine → review → validate → export → serve, each run the same way: kl4a --use sopkb <command>.

That ends with a local web app at http://127.0.0.1:8765: browse sources, review mined knowledge, chat with the in-browser agent. Here's the Inspect bundle screen, browsing by knowledge, on a GLP-1 healthcare SOP set — a proposed item with its evidence and review actions, next to an already-approved one:

Inspect bundle screen, browsing by knowledge: a proposed knowledge item with its evidence and Approve / Reject / Defer / Comment / Edit actions, and an approved item shown locked against further review

And here's the payoff — Agent Studio actually consuming that knowledge to answer a scenario, with its detected concepts, retrieved-context counts, rule evaluations, and evidence shown alongside the answer:

Agent Studio screen showing detected-concept pills, retrieved-context counts, and a rendered answer with rule evaluations, evidence, and caveats

→ Run the full walkthrough on the Quickstart page — every command, expected output at each step, a CLI example of an agent retrieving grounded context, and a second tutorial using a richer DOCX/PDF reference bundle.

Reference bundle

A synthetic, fully worked GLP-1 healthcare example lives at examples/glp1-healthcare. It includes:

  • Multiple SOP-like sources, with DOCX/PDF ingestion
  • Evidence-backed proposed knowledge
  • Persisted human-in-the-loop review states
  • Conflict and freshness reports
  • OKF, Graph JSON, and RDF exports

Where this fits

Free and complete on its own

KL4A handles extraction, evidence grounding, and human review. Free. Local-first. Yours to run anywhere, forever.

Deterministic enforcement, cross-bundle reasoning, audit trails, and governed multi-tenant operation in production are a separate, deliberately out-of-scope concern — a downstream layer this project doesn't try to be. KL4A is a tool for the layer where ontology projects actually fail — this is how your SOPs get ready for that layer, whether or not you ever adopt anything downstream of it. See WHY_OPEN_CORE.md for exactly where that line sits, why, and who maintains this project.

Guides

  • Why This Exists — why KL4A exists and why it's open.
  • Quickstart — step-by-step install and first bundle, with expected output.
  • Desktop UI Guide — the desktop app, screen by screen.
  • MCP Server — connecting an agent to a bundle you've built.
  • Architecture — how the pieces fit together.
  • FAQ — common setup and ingestion problems.

Everything above is sopkb. The other option has its own tab: codekb for a source repository.

Project

KL4A · Apache-2.0 · maintainers & boundary