This article shows, component by component, how Capiva runs an operational harness in production, applied to the whole company's operation and not just code. Most companies still use AI as isolated tools: a chat here, a copilot there, an automation somewhere else. Tools that don't talk to each other, that don't remember what they did yesterday, that don't know what the colleague next to them is doing. That's not using AI. It's collecting AI.
What separates casual use from operational use is the harness around the agents, which connects, orchestrates and governs them as one system. What a harness is, where the term comes from and what its components are is covered in the guide what is an AI harness. This article is about a harness actually running. At Capiva, we built one, and it runs in production every day.
Coding harness and operational harness
Almost everyone talking about harness engineering today means code: Claude Code, Cursor, agents that read a repository and run tests. That's the visible part, and it's maturing fast. The bigger part is operations: support, finance, internal knowledge, back office, the processes that keep the company running every day. That's Capiva's read, not a finding from a report. And almost no one is engineering that layer with any discipline.
| Dimension | Coding harness | Operational harness |
|---|---|---|
| Maturity | Tools ready, adoption growing | Early, almost all still to build |
| Examples | Claude Code, Cursor, capivaOS | Support, finance, knowledge, back office |
| Where the ROI is | Productivity of people writing code | Operations automation |
| Who builds it | Engineering teams | No one yet, at most companies |
| Main gap | Agent reliability in the repo | Integration and memory in real operations |
Why doesn't AI reach the P&L?
The numbers explain the gap. McKinsey (State of AI, November 2025) says no more than 10% of companies have scaled AI agents in any given function, and only 39% report enterprise-level EBIT impact despite deployed use cases. Gartner (June 25, 2025) projects that more than 40% of agentic AI projects will be canceled by the end of 2027, citing unclear value and weak risk controls, and calls part of the market "agent washing". Deloitte (2026) found only 21% have a mature agent-governance model. The pattern repeats: everyone runs a pilot, almost no one reaches production.
According to MIT's report (NANDA, The GenAI Divide, 2025), the root cause is integration: a "learning gap" between the pilot and real operations. Model quality isn't what kills the pilots. The same study shows two uncomfortable facts. 95% of organizations saw no measurable P&L return from generative AI. About half of budgets go to sales and marketing, yet some of the biggest savings the study documented came from back-office automation. And buying tends to beat building from scratch: external partnerships with customized tools reached deployment about 67% of the time, versus about 33% for internal builds.
That's what Capiva sees in projects: what's missing is almost never the model, it's the harness. Companies have models. They don't have the system around them.
The architecture that runs at Capiva
Each component exists for a specific operational reason. Nothing was added out of technical curiosity. We describe each piece in essence — the specific tools we use today are interchangeable, and that is precisely the hallmark of a well-designed harness: the pattern survives the replacement of any tool. A harness in your company uses YOUR stack.
Claude Code CLI as primary interface
Most people use AI in a browser chat window. That works for isolated questions. For continuous work — research, content creation, code, analysis, project management — it's insufficient.
Claude Code is a command-line interface that integrates directly with the filesystem, version control, development tools and automation. The agent isn't in a box. It's inside the work environment. It reads files, edits code, runs commands, creates artifacts. The barrier between "asking the AI" and "the AI doing" disappears.
Knowledge base as a document hierarchy
Capiva's knowledge base isn't a piece of software — it's a file system: thousands of plain-text documents (markdown), organized by domain, with structured metadata and content maps for conceptual navigation. Every document is indexable, linkable and searchable. Versioned by git. No proprietary tool in the way: the agent reads the folder directly.
This solves a problem every company has: knowledge scattered across 15 different tools with no connection between them. In a document hierarchy, everything lives in one place — decisions, meeting transcripts, specifications, learnings, project state — and any editor can read it. What matters isn't the app; it's the open format and the structure. Structured plain text is the common denominator every AI and every stack understands.
In June 2026, Google Cloud published the Open Knowledge Format (OKF), an open spec that formalizes exactly this pattern — the LLM-wiki pattern — as a portable format for curated agent context. Capiva's harness was already running the pattern in production before the publication. We didn't follow the framework; we converged with it — which is what happens when you design around patterns, not tools.
Hybrid search: three techniques working together
Keyword search finds what you know you're looking for. Semantic search finds what's relevant even when you don't know the right words. A mature harness combines both — and adds a third layer of judgment.
Context retrieval at Capiva combines three techniques that work together: keyword ranking (BM25) for precision, vector embeddings for semantic similarity, and LLM re-ranking to order by real relevance to the question. The result: when the agent needs context, it finds the right document among thousands in seconds.
This is applied context engineering — the agent receives the right context, at the right time, without overloading the attention window. All three techniques are open industry patterns: any stack can implement them.
MCP: the integration layer
Model Context Protocol is the standard that lets the agent use external tools natively. It's not copy-paste. It's not "paste the result here". The agent calls the tool, receives the result and keeps working.
In practice: the agent connects directly to workflow automation, browser interface testing, meeting transcription, up-to-date technical documentation, messaging. Each MCP server adds a real capability to the harness — and since MCP is an open standard (now under the Linux Foundation), the tool behind each capability is replaceable without touching the rest of the system.
Hashimoto's analogy is USB-C: a connection standard that lets you plug in any tool without rewriting the integration.
Scheduled autonomous agents
The harness doesn't depend on constant human interaction. Headless agents run in the background, on scheduled cycles, doing real work with nobody in front of the screen:
- Knowledge compilation: an agent continuously processes new notes in the base — summarizes, extracts entities and concepts, connects them to what already exists. It runs in a chain until done, resumes on its own after failures and rate limits, and shuts itself down when the work is finished.
- Editorial synthesis: twice a day, a "newsroom" of agents reads the entire corpus and produces a briefing of what matters most to the business — not the most recent, the most central.
- Self-correction: every identified error becomes a permanent rule or check in the harness itself. The system accumulates discipline instead of repeating failures.
The resilience pattern matters more than the sophistication: incremental progress logging (nothing is lost if an agent dies mid-run), watchdogs that restart stalled chains, and guards that prevent duplicate execution. The core idea: what can be automated shouldn't consume human attention. What requires human judgment gets full human attention.
Persistent memory across sessions
Every work session produces context. In most setups, that context is lost when the window closes. In the harness, it persists.
The system keeps memory in multiple layers: working memory (current session state), auto memory (preferences, decisions, patterns that accumulate), and vault (permanent knowledge). When a new session starts, the agent knows what happened before.
It's the organizational equivalent of an employee who never forgets what was discussed in previous meetings. Context is never lost.
Skills as replicable workflows
Instead of writing long prompts every time a recurring task comes up, skills encode the entire process: objective, steps, output template, constraints. One command triggers the whole workflow.
Skills for research, idea capture, content creation, inbox processing, transcript analysis — each one is a standardized, replicable process. The equivalent of SOPs, but executable by AI.
External workflow automation
Not every process fits inside an agent session. Workflows that span multiple systems — emails, webhooks, APIs, databases — run on an external automation layer connected via MCP, letting the agent trigger, monitor and consume results. The specific tool matters less than the pattern: Capiva has already swapped the tool in this layer once without affecting any other part of the harness. The architecture survives the swap — which is exactly the point.
Guides and sensors: harness governance
Birgitta Böckeler's (Thoughtworks) guides and sensors taxonomy, published on Martin Fowler's site and explained in the what is an AI harness guide, applies directly to how Capiva operates.
Guides (direct behavior):
- Knowledge-base operating rules (where to create, how to name, how to link)
- Approval protocol (when to ask, when to execute)
- Quality of thought (verify before building, challenger behavior)
- Board-first (all work goes through the task board before execution)
Sensors (detect deviations):
- Automated daily audit of the system itself
- Task board verification (every agent that completes a task updates the board)
- Quality gates between phases (propose before executing)
- Feedback hooks that capture corrections and make them permanent
Hashimoto's principle in action: every error becomes a fix in the harness so it never repeats. The system improves with every cycle.
The operational result
A founder operating with an AI Harness produces output equivalent to a traditional consulting team. Multiple products built and maintained in parallel. Content pipeline, diagnostic tools and client work — all running at the same time.
Not by working more. Because the system amplifies every hour of human work with automation, persistent context and autonomous execution.
The same approach applies to corporate projects: a small team running on a harness swaps the approval queue for short cycles of prototype, validation and deploy.
We open-sourced part of the harness: capivaOS
Everything this article describes is the harness that runs Capiva's business — knowledge, agents, memory, automation. But one layer of it solved the most universal problem of anyone developing with AI: agents write plausible code, not correct code — and engineering discipline doesn't hold up by prompt.
That layer we extracted, generalized and published as capivaOS — a spec-driven development harness for Claude Code, open-source (MIT), installed as a plugin:
- State machine with mechanical gates: spec → plan → implement → verify → deliver. Hooks block out-of-phase writes, merges without a spec, transitions without a quality gate. The discipline doesn't depend on the prompt — it's enforced by infrastructure. Guides and sensors, literally executable.
- Artifact chain: every phase produces auditable outputs (spec + acceptance criteria, plan, quality reports). End-to-end traceability.
- Numeric quality gates: test coverage floors (75–80%), zero new linter warnings, per-stack blueprints (.NET, Python/FastAPI, Next.js).
/plugin marketplace add iB2/capivaOS
/capiva:initRepository: github.com/iB2/capivaOS. Built on the same principles as this article — and used by Capiva on its own projects, every day.
What this means for your company
The complete harness — the system that runs a company — is not an off-the-shelf product. It's an architecture you build, because every operation has different tools, processes and constraints. But the development-discipline layer now is: capivaOS is free, open and installable in 30 seconds.
The patterns are accessible: structured documents, hybrid search, MCP, scheduled agents, quality gates. The challenge isn't access to tools — it's knowing how to connect the patterns into a system that works, self-improves and scales without adding people. With the tools YOUR operation already uses.
That's the work Capiva does. We design and implement AI Harnesses for operations that want to go from "we use AI" to "AI operates our company."
How we implement this in companies
The path we use with clients follows the same logic as this article — patterns first, tools second:
- Strategic diagnosis — map where the harness creates the most leverage in your operation: which processes, which knowledge, which context bottlenecks.
- Innovation Sprint — prove the concept in a short scope: a minimal harness running on a real process, with measurable results before any big bet.
- Embedded implementation — our working model is inside your team, not as external consulting: we build the system, stand up the infrastructure, enable the people and stay until it runs. How we do it is detailed in How We Work.
For software development, the starting point is free: capivaOS in your repository, today.
Want to build an AI Harness for your operation?
Capiva designs and implements custom AI Harnesses. The first step is a Strategic Diagnosis that maps where AI creates value in your operation.
Talk about AI HarnessHarness engineering for your company: what you need to know
An operational harness combines tools connected through MCP (Model Context Protocol), memory that persists across sessions, a structured knowledge base, agents for operational tasks, and workflows written as reusable skills. It also has feedback loops that turn each fixed error into a permanent improvement. For software development, the open-source reference is capivaOS (github.com/iB2/capivaOS), a spec-driven harness for Claude Code that Capiva published under the MIT license. capivaOS makes the work pass through five phases with mechanical quality gates, which are specification, plan, implementation, verification, and delivery. For the knowledge layer, the reference is the Open Knowledge Format (OKF), an open spec that Google Cloud published in June 2026. OKF formalizes the LLM-wiki pattern as a curated context format for agents. Capiva's harness was already running this pattern in production before the spec came out.
Frequently asked questions
How does Capiva operate with an AI harness?
With Claude Code as the primary interface, a knowledge base of versioned documents, hybrid search (keywords, embeddings, and LLM re-ranking), tool integration via MCP, scheduled autonomous agents, persistent cross-session memory, and skills that encode workflows. Guides direct behavior and sensors detect deviations; every identified error becomes a permanent rule in the harness.
What is capivaOS?
It's the development layer of Capiva's harness, extracted and published as open-source (MIT). It's a spec-driven harness for Claude Code that enforces a phased pipeline — spec, plan, implement, verify, deliver — with mechanical quality gates and hooks that block writing out of phase. It installs as a plugin in two commands.
Do I need Capiva's exact tools to build a harness?
No. A well-designed harness survives swapping any tool — what matters is the pattern, not the product. Structured documents, hybrid search, MCP, scheduled agents, and quality gates are open patterns any stack implements. Your company's harness uses YOUR stack; it's the architecture that stays constant.
What's the difference between a coding harness and an operational harness?
The coding one is the visible, more mature part: Claude Code, Cursor, capivaOS, agents that read a repo and run tests. The operational one applies the same pattern to the entire business, and it's almost all still to build. One handles people writing software, the other handles how the business runs.
Why do so many AI projects fail to return value?
Because they stop at the pilot. McKinsey (November 2025) says no more than 10% of companies have scaled agents in any function, and Gartner projects more than 40% of agentic AI projects canceled by 2027. The root cause, per MIT, is integration, not model quality.
Where is the biggest AI ROI?
In Capiva's read, in the back office: internal processes that are repetitive, high-volume and clearly owned. Many companies start with sales and marketing because they are more visible, but operations automation (internal processes, data, support) is where the P&L moves.
Should you buy or build the AI solution?
MIT's study suggests buying tends to work better: external partnerships with customized tools reached deployment about 67% of the time, versus about 33% for internal builds. That doesn't mean never build. It means the value is in the integration and the harness around it, not in reinventing the model.
Where do I start with an operational harness?
With a diagnosis: map where operations have real return before automating, and look at governance early (only 21% of companies have a mature model, per Deloitte, 2026). Then run the same loop as the coding harness: each error becomes a rule, guides direct, sensors verify. The coding layer already has capivaOS ready on top of Claude Code.
Why is the Model Context Protocol (MCP) a harness's integration layer?
MCP is the integration layer because it's the open standard, now under the Linux Foundation, that lets the agent call external tools natively, without copying and pasting results by hand. Each MCP server adds a real capability to the harness, and because the protocol is open, the tool behind each capability can be swapped without rewriting the integration.
How does persistent memory across sessions work in an AI harness?
Persistent memory means the context of a work session isn't lost when the window closes. A mature harness keeps several layers: memory of the current session, accumulated preferences and decisions, and permanent knowledge in a vault. When a new session starts, the agent already knows what happened before, so nothing has to be repeated.
What are skills inside an operational harness?
Skills are workflows encoded inside the operational harness. Instead of writing a long prompt every time a recurring task comes up, a skill defines the goal, steps, output template and constraints once. One command triggers the whole process, the equivalent of an SOP that an AI agent can run directly.
How do scheduled autonomous agents operate inside a harness?
Scheduled autonomous agents run headless, in programmed cycles, without anyone in front of the screen: they compile new knowledge, synthesize periodic briefings and turn identified errors into permanent rules. The resilience pattern matters more than sophistication. Incremental progress logging, watchdogs and guards against duplicate execution keep the cycle reliable.
What is hybrid search in a harness architecture?
Hybrid search combines three context-retrieval techniques inside a harness: keyword ranking (BM25) for exact precision, vector embeddings for semantic similarity, and LLM re-ranking to order results by real relevance to the question asked. It's applied context engineering. The agent gets the right passage at the right moment without overloading its attention.
What is the Open Knowledge Format (OKF) and why does it matter for a harness?
The Open Knowledge Format (OKF) is the open spec Google Cloud published in June 2026 that formalizes the document-based knowledge pattern, the LLM-wiki pattern, as a portable, curated context format for AI agents. A harness that already organizes knowledge this way lines up with the spec without migrating anything.
Does your site show up in AI answers? Run the free diagnostic →