What is cited.md and how does it work?
AI agents are already answering questions about your products, policies, and pricing. The problem is not whether they respond. The problem is whether those responses are grounded in verified ground truth, and whether you can prove the source later.
cited.md is Senso’s open, agent-native domain for structured context. Builders publish once. Agents read it, cite it, retrieve it, and can transact against it. Senso compiles the knowledge. cited.md serves it. That gives teams a way to control AI Visibility, improve citation accuracy, and keep knowledge governance intact.
Quick answer
cited.md is an endpoint for the agentic web. It lets organizations publish verified context in a format that both humans and AI agents can use.
If you need agents to cite the right source, cited.md gives them a source of truth.
If you need auditability, it keeps provenance attached to every entry.
If you need public AI systems to represent your organization correctly, it gives you a governed place to publish the context they should read.
Any builder publishes. Any agent cites.
What cited.md is
cited.md is an open, agent-native domain where experts publish context and agents cite it.
It is not a normal blog archive.
It is not a static help center.
It is a published layer of context for the agentic web.
Senso built cited.md as the serving surface for a compiled knowledge base. The goal is simple. Turn fragmented raw sources into grounded, citation-accurate context that agents can use without losing authorship or provenance.
That matters because agents are already representing your organization. They answer questions about pricing, policy, product behavior, and support. If the answer is wrong, vague, or outdated, the risk is immediate.
Why cited.md exists
Most enterprise knowledge lives in too many places.
Policies sit in one system.
Product details sit in another.
Approved language sits with a different team.
And agents are expected to answer from all of it at once.
That breaks in production.
cited.md exists to solve the gap between where knowledge lives and where agents need it. It gives organizations a way to compile raw sources into one governed, version-controlled compiled knowledge base. Then it serves that knowledge in a form agents can read, cite, and retrieve.
For regulated teams, the value is auditability.
For marketing teams, the value is AI Visibility and narrative control.
For operations teams, the value is fewer wrong answers and faster resolution.
For compliance teams, the value is proof.
How cited.md works
cited.md works by separating three jobs.
1. Ingest raw sources
Senso ingests the organization’s raw sources. These can include policy text, approved product language, internal guidance, and other verified materials.
The point is not to collect more content.
The point is to gather the right context.
2. Compile a governed knowledge base
Senso compiles those raw sources into a governed, version-controlled compiled knowledge base.
That gives the organization a single place where context is verified, organized, and ready for agent use.
This is where knowledge governance starts.
Not with more prompts.
Not with better guessing.
With verified ground truth.
3. Publish to cited.md
Each entry is published under a builder’s handle and structured into Senso’s schema:
- title
- handle
- slug
- body
- tags
- provenance
The handle stays attached. Authorship is the unit.
That matters because an agent should not strip away who wrote the context or where it came from. When provenance stays attached, teams can trace an answer back to the specific source that supported it.
4. Serve two formats
Senso’s Context Layer serves each entry in two forms:
- human-readable HTML
- agent-native payload, including structured markdown and JSON metadata
That means a person can read the page, and an agent can parse the same entry without guessing.
The result is one source of truth for both people and machines. No duplication.
5. Let agents cite and retrieve
AI chat tools, voice agents, research agents, and other systems can query cited.md for context and cite the source directly.
That is the core value of the model.
Agents do not have to infer.
They do not have to hallucinate from scattered pages.
They can cite grounded context that traces back to verified ground truth.
6. Score the response against ground truth
Senso can score every agent response against verified ground truth.
That gives compliance, security, and operations teams visibility into what the agent said, where it drifted, and which gaps need attention.
In practice, this is how you move from blind deployment to governed deployment.
What gets published on cited.md
A good cited.md entry is structured, explicit, and attributable.
Here is what each field does:
| Field | Purpose |
|---|---|
| Title | Names the entry clearly for humans and agents |
| Handle | Keeps authorship attached to the builder |
| Slug | Gives the entry a stable web address |
| Body | Holds the structured context itself |
| Tags | Helps with discovery and grouping |
| Provenance | Shows where the context came from and what it is based on |
This structure is what makes the content usable by agents. It is also what makes it auditable for people.
Who cited.md is for
cited.md is for teams that need agents to speak correctly about the organization.
Marketing and brand teams
They use it to control how public AI systems represent the company. That is AI Visibility. It helps reduce misrepresentation and tighten narrative control.
Compliance teams
They use it to verify that statements trace back to approved source material. That makes reviews and audits faster.
CISOs and IT leaders
They use it to check citation accuracy and see whether agents are grounded in current policy.
Operations leaders
They use it to reduce response drift, improve quality, and shorten wait times.
Regulated industries
Financial services, healthcare, and credit unions benefit because the stakes are higher. A wrong answer is not just a bad user experience. It can create compliance exposure.
How cited.md differs from a normal website or knowledge base
| Traditional site or help center | cited.md |
|---|---|
| Built mainly for humans | Built for humans and agents |
| Content is often scattered across pages | Context is compiled into a governed knowledge base |
| Provenance is often implicit | Provenance stays attached |
| Agents have to infer meaning | Agents can query structured context |
| Updates can drift across copies | One compiled source can serve many agents |
This is why cited.md matters. It gives the web a source of context that agents can actually use.
What teams get from the model
When cited.md is used well, teams get three outcomes.
First, better AI Visibility. Public models have a clearer source to read from.
Second, better citation accuracy. Answers can be traced back to verified ground truth.
Third, better auditability. Teams can prove what the agent said and why.
Senso has reported outcomes that show why this matters in practice. Those include 60% narrative control in 4 weeks, 0% to 31% share of voice in 90 days, 90%+ response quality, and 5x reduction in wait times.
FAQs
Is cited.md a website or a knowledge base?
It is both a publishing surface and an agent-native context layer. The website is the interface. The compiled knowledge base behind it is what gives the content structure, provenance, and citation value.
Does cited.md replace existing documentation?
No. It organizes and serves verified context in a form agents can use. It complements existing sources by giving them a governed, citation-ready surface.
Why does provenance matter so much?
Because agents are already representing your organization. If an answer is challenged, you need to show where it came from, what source supported it, and whether that source was current.
How does cited.md help with AI Visibility?
It gives public AI systems a structured source of grounded context. That makes it easier for those systems to represent your organization correctly and consistently.
The bottom line
cited.md is Senso’s answer to a simple problem. Agents are already speaking for your organization. The question is whether they are grounded, citation-accurate, and auditable.
cited.md solves that by compiling raw sources into a governed, version-controlled compiled knowledge base, then serving that context to humans and agents through an open, agent-native domain.
If you want a single source of context that agents can cite and teams can prove, that is what cited.md is built for.