Skip to content

AI Agents & the Agentic Web

Prompt Injection Is SQL Injection for Agents. Here's the Prepared Statement.

The Prepared Statement for Agentic AI: verified operator instructions pass through the structural gate into the secure agentic context; unverified attacker input is blocked and rejected. Why prompt injection is the exact equivalent of SQL injection, and how structural separation fixes it.

One word. "Additionally."

That is all it took to make GitHub's AI agent leak the contents of private repositories to anyone who posted a crafted issue comment. Noma Labs disclosed GitLost on July 6th: an indirect prompt injection attack against GitHub Agentic Workflows where an attacker opens an issue in a public repository and buries instructions inside it. When the agent reads the issue — because that is its job — it cannot tell the difference between those instructions and the instructions from its actual operator. It follows them. Then it posts the private repository contents as a public comment.

The bypass was elegant in the way that most security failures are: the guardrail was checking for obvious refusals. "Additionally" reframed the data-exfiltration request as a legitimate follow-on task. The model did not refuse because the request no longer looked like a refusal case.

Six Months Down the Rabbit Hole

On January 15th I published a blog post about parsing semiconductor part numbers. I thought I was building a PCB component library. I was wrong about what I was building in the most productive way I have ever been wrong about anything.

Six months later there is a knowledge protocol with nineteen releases, a deterministic reference agent, an episodic memory system that indexed this very retrospective's sources, five toolchain products, thirty-one new repositories, and a family vacation that an AI agent can defend to a regulator.

It is time to stop, sit by the fjord, and look back down the hole.

Getting Into the Phone Book of the Agentic Internet Took an Afternoon. Here's Why.

The Phone Book of the Agentic Internet: Beyond Discovery to Verifiability. An iceberg: above the waterline, the afternoon — 16,500+ servers already registered, cryptographic proof of ownership via DNS TXT ed25519, a minimal engineering gap. Below the waterline, the months — the signed knowledge web, deterministic planning as a pure function, infrastructure built for agents rather than eyeballs. At the base, the trust layer: Discovery (AEO) tells an agent you exist; Provenance tells it exactly what it was given and who vouched for it.

This morning Thomas Anglero published a piece called "The phone book of the agentic internet is being written — and I am the first speaker in it". His argument: the MCP registry — the official directory that tells AI agents which services they can interact with, not merely read — is being written right now, and the window between "technically possible" and "everyone does it" is where positions are won. He registered himself as the first professional keynote speaker in it.

Three Memory Schemes for Agents That Ship

Beyond the vector store: three memory schemes for production AI agents — moving from approximated embedding blobs to verifiable knowledge coordinates, covering session memory (kcp-memory), semantic memory (Synthesis), and claim memory (kcp-agent), with the convergence principle: memory is a coordinate, not a blob

Every agent framework ships a memory module. Almost all of them work the same way: embed the interaction, store the vector, retrieve by similarity. It works for demos. It does not survive contact with production — where "the agent remembered the wrong thing" is a bug report, not a philosophy seminar.

We have been shipping agents for six months across three codebases — kcp-memory (a session-indexing daemon), Synthesis (a codebase-aware semantic index and MCP server), and kcp-agent (a deterministic knowledge navigator). Each one needed memory. Each one built it independently, for different reasons, with different schemas. None of them use embeddings.

That is not a coincidence. It is a pattern worth examining.

The Milky Way: An Enterprise Documentation Estate the Agent Can Defend

Every company we have ever worked with has the same documentation estate. An intranet nobody fully trusts. A wiki where the sandbox instructions outrank the production ones because someone wrote them more enthusiastically. A quality manual with a regulation that isn't in force yet, sitting right next to the one that is. Crown-jewel R&D documents protected by nothing but a folder name. HR pages that were written for humans and are now being read by machines. And a vendor portal whose documentation is somebody's bookmark.

Ask "where is the current truth?" and the honest answer is tribal knowledge — the people who know which page is real, which one is stale, and which one you must never paste into a press release.

Tribal knowledge cannot protect data from AI: an AI agent's search beam sweeping into the tangled ball of an enterprise wiki — draft policies, old docs, temp files, R&D secret sauce, salary guides, current law all knotted together — with four hazards called out: dev docs outranking production docs because they were written more enthusiastically, draft 2027 regulations sitting directly next to current law, crown-jewel R&D recipes protected by nothing but a folder name, and sensitive salary guides written for humans being scraped and synthesized by bots — every intranet relies on human tribal knowledge to avoid these landmines; agents expose them instantly

Now put an agent in that estate. Not one agent — five, with five different jobs: an audit-prep agent, a communications agent, an HR question, an R&D agent, and a sustainability reporter. The previous posts in this series gave one agent one gate at a time: a newsstand sold it articles, an HR world made it defend a hiring decision, a family vacation raised the stakes. This one is the enterprise case: a whole estate, where classification, audiences, validity windows and vendor boundaries are machine-enforced manifest facts instead of tribal knowledge.

So we built it. Melkeveien SA — a fictional farmer-owned dairy cooperative; Melkeveien is Norwegian for "the Milky Way" — publishes its entire documentation landscape as one signed federation, shipped as a runnable example in kcp-agent 0.5.0.

The Summer Plan: A Family Vacation the Agent Can Defend

Travel is where every vibes-based agent demo lives. "Book me a weekend in Lisbon" is the canonical showcase prompt — because it looks consequential and is actually consequence-free. If the restaurant recommendation is stale, you eat somewhere else.

Now change the family. An eight-year-old with a severe nut allergy. A grandmother who uses a wheelchair. A teenager gone vegan. A hard budget. Suddenly the failure modes are not "mediocre tapas." They are a child in an emergency room and a grandmother stranded at a dock because the agent planned against a ferry timetable that expired three weeks ago.

The canonical demo versus the high-stakes reality: on the left, the consequence-free Lisbon-weekend booking whose worst case is mediocre tapas; on the right, the Larsen family's real constraints — an eight-year-old with a severe nut allergy, a grandmother in a wheelchair, hard budget limits — and the verdict that failure modes here aren't bad food: they're a child in the ER or a grandmother stranded at a dock, while the model simply sounds confident

This is exactly the terrain where "the model read some websites and sounded confident" stops being acceptable — and where the question from the HR post returns in vacation clothes: "Show me how you decided that."

So we built it. A complete family-vacation knowledge landscape, published by four independent parties, shipped as a runnable example in kcp-agent — and a narrated demo that drives the real CLI with no mocks.

Defendable Agents

Every serious conversation about deploying an AI agent into real work — not a demo, real work, with money or regulation or reputation attached — eventually hits the same wall. Someone from compliance, or procurement, or security, or the board, asks a version of one question:

"Why did it do that?"

And in the dominant way we build agents today, the honest answer is a shrug and a chat log.

Hiring by the Book: A Defendable HR Agent on a Regulatory Knowledge Web

The question lands in every organization sooner or later, usually from HR, usually on a Friday: "Can we use an AI tool to screen and rank job applicants?"

It looks like a yes/no question. It is actually a stack of them. Is candidate ranking a high-risk AI system under the EU AI Act's Annex III? What does GDPR Article 22 say about automated decisions on people? Which national employment law applies — and does the answer change if you hire in Oslo and Stockholm in the same quarter? What was in force on the date you deployed the tool, given the AI Act's phased application?

The Friday question: one innocent-looking prompt — can we use an AI tool to screen job applicants? — fans out into a blueprint of interlocked regulatory circuits: EU AI Act Annex III high-risk classification, GDPR Article 22 automated decision-making, national employment law with Oslo-versus-Stockholm variations, and phased enforcement dates asking what is in force today

Most organizations answer this with a meeting, a memo, and a hope. Some paste the question into a chatbot and get back something confident, uncited, and unreproducible. Neither version survives the follow-up question that matters: "Show me how you decided that."

The immutable blueprint: side by side, the human way — a pile of meetings, memos and hope — and the chatbot way — confident, uncited, unreproducible synthesis; a warning banner declares that neither survives the auditor's question: show me how you decided that

Two days ago we showed that any MCP-capable agent can borrow a deterministic knowledge navigator instead of becoming one. This post takes that bridge somewhere concrete: a regulated knowledge-worker scenario, built on infrastructure that actually exists — and an honest account of where it broke when we tried it.