The Shrink Knowledge Graph.

A knowledge graph is a way of representing connections between things rather than just the things themselves. Instead of a list of articles, you have a structure where every concept connects to other concepts in specific ways. Symptom connects to condition. Condition connects to treatment. Treatment connects to medication. Medication connects to the research evidence. Research connects to the lived experience. Lived experience connects to the practical tools. The tools connect to the care option. The care option connects back to the symptom. The whole thing is one structure.

The Shrink Network is built around a knowledge graph like this. Every entry on every editorial site has a network_topics tag that places it in the graph. Every entry has a Network Compass at the top showing where the entry sits relative to the other sites. Every entry has a Knowledge Path at the bottom showing a directed walk through related entries. Every entry has a Continue Learning block routing the reader to the other sites that cover the same topic. The Network Clusters at shrinkopedia.com/clusters/ are the explicit cross network landing pages built directly on top of the graph.

Why does this matter? Because competitors can copy a page. They can copy a glossary. They can copy a medication entry. What's substantially harder to copy is hundreds of interconnected relationships across multiple branded sites that all reinforce one another. The relationships are the moat. The brands are the moat. The editorial standards consistency end to end is the moat. The knowledge graph is the structural reason why this network is something other than a collection of sites.

If you'd rather jump straight into the encyclopedia's directory pages instead of reading the taxonomy essay first, Shrinkopedia organizes its reference material into five parallel entry points: Conditions (diagnostic entries aligned to DSM-5-TR and ICD-11), Treatments (psychotherapies, medications, and adjuncts), Terms (reference-level explainers on individual concepts), Library (one-page explainers on frameworks and skills), and Symptoms (what the experience looks like before it's a diagnosis). Every graph node in the network resolves to one of these five directory shapes.

Below is the conceptual map of what we connect, how we connect it, and why this approach is what mental health information on the internet should have looked like all along.

What the network connects.

The graph spans seven categories of mental health knowledge. Each category lives primarily on one or two sites in the network, and each category cross links to the others through the structural components on every entry.

• Symptoms. What you actually experience. Racing heart, intrusive thoughts, anhedonia, brain fog, hopelessness. Primary site: Shrinkopedia. Practical layer: AnxietyResource and DepressionResource.
• Conditions. The clinical categories. Generalized anxiety disorder, panic disorder, major depressive disorder, OCD, PTSD, bipolar disorder. Primary site: Shrinkopedia. Practical layer: AnxietyResource and DepressionResource.
• Treatments. The approaches with evidence. CBT, exposure therapy, behavioral activation, ERP, EMDR. Primary site: Shrinkopedia. Evidence layer: AnxietyResearch.
• Medications. The pharmacological options. SSRIs, SNRIs, mood stabilizers, antipsychotics, stimulants, sleep medications. Primary site: PsychiatryRx.
• Research. What the trials actually say. Primary site: AnxietyResearch.
• Tools and Patterns. The skill building and pattern recognition work. Primary sites: shrinQ, Unstuck, QuitScrolling.
• Care. The clinical service. Primary site: shrinkMD.

How the connections work.

Every editorial entry on every site uses the same network_topics taxonomy. A symptom entry tagged panic connects to a condition entry tagged panic connects to a treatment entry tagged panic connects to a medication entry tagged panic. The taxonomy is the connective tissue.

Network components on every entry do the connecting visually. The Network Compass at the top shows where the entry sits relative to the other sites in the network. The Knowledge Path at the bottom shows a five step walk through related entries. The Continue Learning Across The Network block shows the four other network properties most relevant to what the reader just read. None of this is a generic footer or sidebar; it's content aware, calibrated to the specific entry's topic.

The Network Clusters at shrinkopedia.com/clusters/ are the explicit cross network landing pages. Each cluster is a single topic (Generalized Anxiety Disorder, Panic Disorder, Major Depressive Disorder, OCD, Social Anxiety Disorder, Postpartum Depression, and growing) covered from every angle by every site in the network, with all the angles surfaced on a single page. The clusters are the strongest single demonstration of what a knowledge graph based ecosystem can do that a collection of separate sites can't.

Why this approach matters.

Mental health information on the internet is usually fragmented. A search for "Lexapro side effects" lands on a medication site. A search for "what's wrong with me" lands on a symptom checker. A search for "how do I find a therapist" lands on a directory. Each search is treated as a separate problem with a separate destination. The relationships between the searches are invisible to the reader.

But mental health doesn't work in fragments. Someone searching for "Lexapro side effects" also probably wants to understand what the medication is treating, what the condition feels like day to day, what other treatments exist, when to consider getting evaluated, and how to think about whether the medication is actually working. A single destination that covers all of those questions reflects how mental health understanding actually develops.

The Shrink Network is structured around that reality. The reader who arrives at PsychiatryRx with a Lexapro question is routed to Shrinkopedia for the clinical context, to DepressionResource for the practical layer of what depression looks like, to shrinkMD for the care option, and back to PsychiatryRx for the medication specifics. The routing isn't generic. It's calibrated by content. The reader doesn't have to know which sites exist or what each one does; the network does that for them.

What's possible.

The Shrink Network is in active build. The editorial infrastructure (the Compass, Knowledge Path, Trust Block, cross cluster links, the Maps on AR and DR, the Hubs on Shrinkopedia, the clusters on Shrinkopedia) is in place across the four core editorial sites. The smaller sites in the network (AnxietyResearch, shrinQ, Unstuck, QuitScrolling, shariqrefai.com, shrinkiatry, Shrinktionary, shrinkDaily) are being brought to the same standard over time.

What's possible at scale is something the internet doesn't yet have: a structured, branded, clinically reviewed mental health knowledge ecosystem that treats mental health the way clinicians actually think about it. Not as fragmented searches with fragmented answers, but as a connected territory you can walk through from any direction. That's what we're building.

If you're new to the network, the Start Here page routes you based on what brought you in. The Network Map explains the four layers.

Open Start Here → Open the Network Map →

Named systems in the network.

The Shrink Network runs on a set of proprietary systems, each one owned by a specific site. Naming them makes the architecture visible. It also makes it easier to say what belongs where.

Term Web

Shrinktionary's visual map of how mental health terms relate. Click a term and see the neighboring terms, the parent category, and every place the concept shows up across the network. Lives at shrinktionary.com/term-web.

Concept Graph

The machine-readable spine of the network. Shrinktionary publishes /concepts.json as the canonical concept ID list; every other site references those slugs. It's the reason "one shared graph" is a fact, not a claim. Standard at editorial standards.

Pathways

The reader's guided walks across the network, from the first plain-language question up through practical tools and, if it makes sense, into care. Ten pathways live on /pathways/. Each one is a sequence of sites, in order, so the reader doesn't have to know which property does what. ShrinkDaily implements them in-page as "Overthinking Pathway, Step 2 of 5" with cross-site previous and next links.

Psychiatry Operating Room

shrinkiatry's flagship: twelve rooms mapping how the profession actually works, from residency to reimbursement to the ethics of confidentiality. Reader-facing counterpart is the Decoder. Lives at shrinkiatry.com/psychiatry-operating-room.

Daily Concept System

shrinkDaily's cadence: one mental health concept a day, taught as a lesson, referencing a canonical Shrinktionary slug. Realms, pathways, and frameworks organize the concepts over time. Lives at shrinkdaily.com.

Knowledge Path, Compass, Trust Block

Editorial infrastructure shared across the four core editorial sites (Shrinkopedia, AnxietyResource, DepressionResource, PsychiatryRx). Knowledge Path is the read-order for a topic. Compass is the top-of-page reader router. Trust Block names the reviewer and last-reviewed date on every page. Editorial standards for the network are at /editorial-standards/.

Toolkit Library

The canonical index of every downloadable asset in the network. Pulled live from each site's /downloads.json. Lives at shrinknetwork.com/toolkit.

Network Registry

The registry that lets any site find every other site in the network. Machine-readable at /network.json. Every site publishes its own /index.json. Together, they're what makes the ecosystem an ecosystem rather than a directory.

Concepts flow across layers in a predictable order: language (Shrinktionary defines the word) → understand (Shrinkopedia explains it) → learn (shrinkDaily teaches it one lesson at a time) → condition (AnxietyResource, DepressionResource put it in daily context) → evidence (AnxietyResearch summarizes the research) → medication (PsychiatryRx explains the drug) → apply (shrinQ, Unstuck, QuitScrolling turn it into practice) → care (shrinkMD provides the clinical service) → profession (shrinkiatry and shariqrefai.com map the field). That's the flow. Every reader can enter at any step.

The live graph.

Below is the graph as it exists right now, pulled from the network registry at /network.json. Each site publishes its own /index.json, and the sections below are populated from that data. Two views. The first is site-first, one card per property. The second is concept-first, one card per concept, listing every page in the network that carries that concept ID.