Research brief · AI agents · October 2026 · four AI models, Anthropic, OWASP, Cloudflare, arXiv

The agent-readable
second brain

Every company that puts AI agents to work needs one place those agents read from: the procedures, rules, decisions, product facts and approved answers. We call it the second brain. We asked four AI models, Grok 4.7, GPT-5.6, Claude Opus 5.5 and Gemini 3.1 Pro, how a business should build one at five company sizes, and to tear apart the design under review. Here is where they agree, where they split, and what we would build. Research date: 14 October 2026.

See it by company size
67%Fewer retrieval failures with hybrid search plus reranking (Anthropic)
39×More tokens per answer for a file-browsing agent than BM25 search (arXiv 2026)
#1Prompt injection's rank in the OWASP LLM Top 10, via files and tools
4 / 4Models that say files in git, not the database, should be the source of truth
Tall wooden library shelves full of books
The shape · four layers every size shares

Files, index, live data, gate

All four models converge on the same skeleton. People write and approve knowledge as plain files. Agents search an index built from those files, which can be thrown away and rebuilt. Bookings, invoices and stock stay in the systems that own them and are asked for at the moment of need. And nothing reaches the brain, or the customer, without passing a human and a test. What changes with company size is how much machinery sits around each layer.

Layer 1 · The canon

Files in git

Source

Markdown with YAML frontmatter, one topic per note, organised by kind of knowledge: entities, procedures, rules, decisions, FAQ, glossary. Owned, versioned, portable to any model.

Layer 2 · The index

Rebuildable search

Hybrid

Keyword (BM25) plus vector search, built per audience, rebuilt from git in CI. If it is lost, it is rebuilt in minutes. Small companies may not need it at all yet.

Layer 3 · Live systems

Ask, don't copy

MCP

CRM, ERP and PMS data is queried through narrow, read-only MCP tools with their own credentials. A price written into prose is a future wrong answer.

Layer 4 · The gate

Approve and test

Evals

Agents propose changes, a named owner approves, and a known-answer test set runs before anything goes live. All four models say this is where projects live or die.

Why hybrid search, and why not only file reading

Top-20 retrieval failure rate

Cross-check · Grok 4.7, GPT-5.6, Claude Opus 5.5, Gemini 3.1 Pro

Where they agree

  • All 4
    Files are the source, the index is disposable

    Knowledge lives as markdown in git. The vector or search index is a derived artifact you can rebuild, which also lets you swap embedding models without migrating knowledge.

  • All 4
    Live data stays where it lives

    Prices, bookings, stock and invoices are read at task time from the system of record, through scoped tools, not copied into notes or the vector store.

  • All 4
    Agents propose, humans approve

    A poisoned ticket that can commit to the handbook is a permanent injection; one that can only open a pull request is a bad draft. Every model kept this rule.

  • All 4
    Customer and staff knowledge are split physically

    A customer bot should have no credential that reaches the confidential index. Separate indexes beat a metadata filter in one index.

  • 3 of 4
    Hybrid beats vector-only

    Grok, GPT and Claude insist the first index include keyword search, for SKUs, names, law articles and Greek inflections. Gemini puts hybrid only at mid-market.

  • 3 of 4
    MCP is plumbing, not security

    The protocol standardises access, but identity, scopes and consent must be enforced by the host and server. Gemini treats MCP more as the answer than the risk.

  • 3 of 4
    Graph RAG is not a default

    Grok, GPT and Claude say add a graph only after evals show multi-hop failures; links and entity tables give most of the value. Gemini sees it as necessary from mid-market.

  • 3 of 4
    Buy search for the long tail at scale

    From roughly 100 people, suite search (Copilot, Glean, Notion) handles "find any document I can see", beside a smaller curated canon for agents that answer customers or act.

Where they disagree

When does a company need a search index?

Read the files directly as long as possible, or index early? Placed by the size at which each model adds an index.

Index from day oneFiles only until it hurts

How many test questions are enough?

The reference design suggests 20 to 30 known-answer questions on every change. Placed by how large each model wants the set.

20–30 is fine100–300 or more

How soon does a knowledge graph pay off?

GraphRAG for complex questions, or links and tables first? Placed by how central each model makes the graph.

Rarely, lateCore from mid-market

Pin positions are House of 47's reading of each report, not scores the models gave. "Synthesis" is our own view, which is opinion.

By company size · what to build, October 2026

Five sizes, one skeleton

The synthesis below takes the point where at least three models agree, and flags where we chose a side. Architecture and effort are recommendations, so opinion. Prices are third-party list prices with sources, checked by the models on 14 October 2026; they exclude model tokens and staff time, which all four models say dominate the bill.

Solo / micro
1–5 people

Stack. A private git repo of markdown, edited in Obsidian or any editor (Obsidian is free for commercial use; Sync is $4 a user a month). The agent always loads a short _index.md or "how to read this" file, then opens the notes it needs, or loads the whole folder with prompt caching while it is small. No vector database yet.

Structure. Folders by kind of knowledge, a 1–2 sentence summary at the top of each note, and frontmatter with at least id, type, audience, owner, status and review date.

Write path. The founder owns everything. Agents drop proposals into an inbox folder or a pull request. A weekly 30-minute review.

Permissions. A separate public folder for any customer bot, from the first day. Read-only, narrowly scoped keys for live systems.

Evals. 10–50 real questions, including ones the bot must refuse, run by hand after each change.

Cost. About $0–35 a month of infrastructure; days to set up; 1–2 hours a week. A $20–25 seat plan (ChatGPT Business, Claude Team) may be enough if you only need an internal assistant.

Fails when. Personal notes quietly become customer truth, or a price in a note goes stale.

Small
5–20 people

Stack. The same repo with protected branches and code owners (GitHub Team, $4 a user a month). Add a hybrid index when a second audience, a few hundred notes or several agents appear: Postgres full-text plus pgvector (Supabase Pro from $25 a month) or a managed service such as Cloudflare AI Search (billing starts 1 Nov 2026, $0.75 per 1,000 semantic queries after 1,000 free). Serve it through one retrieval MCP server.

Structure. Add a glossary with Greek, unaccented and Greeklish aliases, and keep exact facts in small data tables, not prose.

Write path. An owner and a backup owner per domain. Agents open pull requests; the owner merges. A weekly stale-content queue.

Permissions. A physically separate customer index and credential. The caller never chooses its own audience tag.

Evals. 20–30 questions in CI at minimum, plus refusals, leak tests and one planted injection. Claude wants 80–150.

Cost. About $10–150 a month of infrastructure; days to a few weeks to set up; a few hours a week.

Fails when. Vector-only search misses exact codes, or non-technical staff avoid git and stop editing.

SMB
20–100 people

Stack. Git canon plus Postgres doing keyword and vector search in one place, rebuilt in CI. A central retrieval service, an MCP gateway, and audience-specific tools. Add a reranker only if evals show a gain.

Structure. A schema enforced by a linter: required frontmatter, no retired notes in the index, currency amounts flagged if the number is live elsewhere.

Write path. The bottleneck moves from search to authorship. Domain stewards, risk-based approvals (two approvers for pricing, legal, safety), and a web "suggest a change" form that produces the same pull request, because Obsidian shared vaults cap at 20 collaborators.

Permissions. SSO groups, database roles or row-level security, sensitive domains in separate repos, and end-user OAuth on live-data MCP so a customer agent sees only that customer's booking.

Evals. 50–400 questions by audience, kind and language, with retrieval scored separately from answers.

Cost. About $100–1,000 a month of infrastructure, or about $2,000 a month for 100 Copilot Business seats at $21 list. Weeks to set up; 0.2–1 FTE.

Fails when. "Connect everything" imports drafts and contradictions, or access metadata is lost during chunking.

Mid-market
100–500 people

Stack. Split the job. Buy permission-aware search for the long tail of documents, mail and chat (Microsoft 365 Copilot, Glean, Notion Business at $20 a member). Keep a smaller governed canon for agents that answer customers or act, exposed through MCP behind a gateway that passes the user's identity through.

Structure. Domain canons with stable IDs linked to CRM and ERP master data; derived entity links where multi-hop questions are real.

Write path. A knowledge owner, domain stewards and segregation of duties, wired into existing ticketing.

Permissions. SSO and SCIM, attribute-based access, separate runtimes for customer agents, a DPIA where personal data is involved. Residency is decided per product, not per vendor.

Evals. Per-domain sets built from real traffic and incidents, red-team suites for exfiltration and poisoned documents, quality tracked per language and department.

Cost. Licences dominate: 300 seats at $20–48 is roughly $6,000–14,000 a month. Canon infrastructure is low hundreds to $3,000. About a quarter to roll out.

Fails when. A central team becomes the approval bottleneck, or a suite connector keeps the documents but not the nested permissions.

Enterprise
500+ people

Stack. Enterprise search or an existing cloud search stack for the unmanaged corpus; a curated, schema-enforced canon for policies, products and approved answers; agents that act get the canon and scoped tools, never "search the whole company".

Structure. A governed federation and catalog that preserves records management, legal holds and source permissions. Do not copy the company into markdown.

Write path. Agents draft from tickets and transcripts; stewards and a council approve; content approver, platform operator and auditor are different people.

Permissions. Zero-trust, policy-as-code, workload identity, regional deployment, customer agents in separate accounts with their own subprocessors.

Evals. A continuous programme: stratified golden sets, shadow evaluation, red teams, rollback thresholds.

Cost. 500 Copilot seats at $30 list is $15,000 a month before Microsoft 365 licences. Glean publishes no price; Vendr reports a median contract of $98,890 a year. A standing team of 3–10 FTE.

Fails when. "One enterprise brain" becomes one exfiltration surface, or the pilot ignores deletion, legal hold and group changes.

Side by side · the five sizes

The comparison

One row per decision, one column per size. Scroll the table sideways on mobile.

Solo / micro 1–5Small 5–20SMB 20–100Mid-market 100–500Enterprise 500+
Source of truthPrivate git repo of markdownSame, protected branchesGit canon plus web suggest-a-changeDomain canons; sensitive domains in own reposCurated canon; the rest stays in source systems
How agents readLoad index note, open filesHybrid search, then open fileHybrid search per audience roleCanon via MCP gateway plus bought searchIdentity-aware platform and gateway
Live dataOne read-only MCP toolTwo or three, separate secretsMCP gateway, end-user OAuthPer-agent credentialsPlus human approval on risky writes
Who writesThe founder; agents proposeOwners per folder; agent PRsStewards; risk-based approvalKnowledge owner and stewardsStewards and a council
PermissionsSeparate public folderSeparate customer indexGroups, row-level securitySSO, SCIM, attribute-basedZero-trust, policy-as-code
Eval set opinion10–50, by hand20–150 in CI50–400, by audience and languagePer domain, from tracesContinuous release gates
Infrastructure excl. seats, tokens$0–35 / mo$10–150 / mo$100–1,000 / mo$500–3,000 / moContract-dependent
People opinion1–2 h a week2–6 h a week0.2–1 FTE0.5–3 FTE3–10 FTE

Ranges merge the four reports' planning estimates and are opinion, not quotes. Seat prices that swamp the table if you choose them: Microsoft 365 Copilot Business $21 and Enterprise $30 a user a month, Notion Business $20, Cloudflare Access $7 after 50 free users.

When to move up

  • Trigger
    Files only → files plus index

    One audience's canon passes about 30–50k tokens, agents open more than about five notes per answer, recall drops on paraphrased or Greek questions, or a second audience appears. The audience is the boundary, not the agent count.

  • Trigger
    One index → one per audience

    The first customer-facing agent, or the first content some staff may not see.

  • Trigger
    Obsidian for everyone → curators plus a proposal flow

    More than about ten regular editors, repeated sync conflicts, or non-technical staff avoiding git.

  • Trigger
    Canon only → canon plus bought search

    Staff ask agents about contracts, project folders and mail that will never be curated. Buy suite search; do not ingest the mailbox.

  • Trigger
    Informal → formal audit and DPIA

    The first agent that acts rather than answers, the first regulated domain, or an auditor asking where a prompt is processed.

  • Stay put
    When the eval is green

    If the owner is real and the tests pass, a new product announcement is not a reason to move. Grok's line, and we agree.

Triggers are opinion, but each one is observable. Thresholds come mainly from Claude Opus 5.5 and GPT-5.6.

Critical assessment · the design under review

Right shape, missing contract

The design under review has three layers: people write and approve in an Obsidian vault (markdown and git); PostgreSQL with pgvector, or a Cloudflare stack of R2, AI Search, Vectorize, D1, Durable Objects, AI Gateway and the Agents SDK, serves as a rebuildable index; live data is queried through MCP. Six rules sit on top: files are the source, structure by kind of knowledge, live data stays in place, permissions from day one, agents propose and humans approve, and 20–30 known-answer tests on every change. All four models called the shape right. Grok's summary fits: most of what is good is organisational, most of what is weak is a missing contract underneath a sound diagram.

What is right

  • All 4
    Files first, index second

    "Never the reverse" is, in every report, the most important decision in the design.

  • All 4
    Live data via MCP, read-only

    Keeping upstream secrets in the MCP server, not the agent, is the correct trust boundary.

  • All 4
    Pull-request approval

    The only write path that matches the knowledge-poisoning threat, using a review tool people already know.

  • All 4
    Evals on every change

    Directionally the whole quality programme in one rule.

  • 3 of 4
    A separate index per audience

    "The customer chatbot has no connection to the confidential index" is the strongest security idea in the notes (Claude, GPT, Grok).

  • 2 of 4
    Start with file reading

    Adding the index later matches the evidence on agentic search and long-context decay (Claude, Grok).

What is wrong or missing

  • 4 of 4
    20–30 questions is a smoke test

    With 25 questions, one answer is worth 4 points; Claude computes a 95% interval of about 61–91% around a measured 80%. Add refusals, leaks, injection fixtures and Greek variants, and grow the set.

  • 3 of 4
    The EU residency check fails for the index

    R2, D1 and Durable Objects can be pinned to the EU. Vectorize documents no EU jurisdiction, and Cloudflare's own table marks Vectorize, AI Search and Workers AI incompatible with Regional Services. Use EU-region Postgres for EU-only clients.

  • 3 of 4
    Obsidian is an editor, not the system

    No SSO, no live co-editing, a 20-collaborator cap with owner-level access for all. Non-technical staff need a web form that makes the same pull request.

  • 3 of 4
    pgvector is not hybrid search

    Add Postgres full-text (with the greek stemmer and unaccent), rank fusion and filters. Vectorize has no keyword search; D1 is SQLite with a 10 GB cap.

  • 3 of 4
    Memory and logs sit inside the brain

    Agent memory is unreviewed, per-user and often personal data. Keep it in a separate store, read-only for agents that read untrusted input; it reaches the canon only by pull request.

  • 2 of 4
    No file schema, no freshness

    No owner, status, review_by or supersedes fields; no stale-note report. Maintenance is the failure the notes themselves predict.

  • 2 of 4
    Tested after going live

    Webhook sync means a bad change is live before the score drops. Build a staging index, run evals in the PR, then swap (Claude, GPT).

What we would change

AreaIn the notesChangeRaised by
File contract"Entities, procedures, rules, decisions, FAQ"CommonMark plus YAML with id, type, audience, owner, status, lang, aliases, review_by, supersedes; linted in CIGrok, GPT, Claude
Retrievalpgvector, or Vectorize plus D1Hybrid from the first index; Greek normalisation; rerank only if evals say soGrok, GPT, Claude
AudiencesPermissions added as the company growsPublic/internal split mandatory at every size; audience-first directories that fail closed; sensitive domains in separate reposGrok, Claude
Live data"Never copied", read-only where possibleEnd-user OAuth on MCP; allow labelled caches with TTL for outages (GPT)GPT, Claude
Evals20–30 known answers per changeStart there, grow to 100–300 stratified; score retrieval and answers separately; gate a blue/green index swapAll 4
Residency"Check before the DPA"EU-region Postgres for EU-only clients; name every processor, including the generation model, in the DPAGrok, GPT, Claude
Write pathOne-click PR approvalProvenance in every PR, two approvers on pricing and policy, instruction-text checks, a non-git proposal formAll 4
Answer contractNot in the notesCite note IDs, abstain below a threshold, escalate; disclose AI to customers (EU AI Act Article 50, since 2 Aug 2026)GPT, Claude

The notes' path from files only, to files plus an index, to per-audience indexes with memory and logs, is sound in order. The models' main objection is timing: audience separation, the schema and hybrid search belong at the start, not at a later step. The "when you move on" line (past a few hundred pages, or a second agent) is reasonable; Claude would trigger on a second audience rather than a second agent.

Against the alternatives

ApproachWins whenLoses whenVerdict opinion
Suite copilots and enterprise searchCopilot, Glean, Notion AIStaff ask "what did we say in our own tools?" across many sources; permissions already cleanCustomer-facing bots; seat cost grows with humans, not agents; inherits oversharingBuy beside the canon from about 20–100 people; never make it the brain
Graph RAGGraphRAG, LazyGraphRAGGlobal and multi-hop questions over large corporaLookup questions; full builds are costly and lost to BM25 at scale in one 2026 studyLinks and entity tables first; graph only where evals prove it
Long context"just load the files"A clean handbook that fits, with steady traffic and cachingNoisy corpora; context rot; 150k uncached tokens is about $0.30 a question on Sonnet 5.5Always load a short guide; retrieve the rest
Hosted file search and memoryOpenAI, xAI, Gemini, AnthropicPrototypes and public FAQ bots; $2.50 per 1,000 calls at OpenAI and xAIAs the system of record: vendor coupling, less control, separate residencyFine as a rebuildable index or scratch pad, never the only copy
MCP knowledge serverMany agent runtimes need one read contract: search, fetch, list changesEach server is its own permission and audit surfaceThe right interface for the canon; keep writes on the PR path
Probability of outcome · House of 47 synthesis

Where it goes next

Synthesis estimates from the four models' outlooks, scenario estimates, not forecasts of record. The through-line: vector search becomes a commodity, and identity, policy and review become the product.

Which model said what · one prompt, four AI models · 14 October 2026

Which model said what

We gave the same research prompt to four AI models on 14 October 2026: Grok 4.7, GPT-5.6, Claude Opus 5.5 and Gemini 3.1 Pro. Each wrote its own report on what to build at five company sizes and assessed the same reference architecture. Here is what each one said, where they agree and where they split.

Method: one prompt, run once per model on 14 Oct 2026. The prompt asked for a source link behind every number, a clear line between fact and opinion, and honesty about each model's own maker. We compared the four reports side by side and quote their positions as written, shortened. Where a model's claim conflicts with another's sourced figure, the sourced figure stands. Gemini 3.1 Pro's report is the shortest, lists its citations as titles without links, and names older models (Claude 3.5 Sonnet, Gemini 1.5 Pro) as examples.

QuestionGrok 4.7GPT-5.6Claude Opus 5.5Gemini 3.1 Pro
Solo / microGit repo plus a "how to read" file; index only when it outgrows a careful readMarkdown or existing workspace; direct reading with a manifestLoad the folder with caching, or index note plus grepLong context, a local MCP server, or Notion AI
SmallHybrid search, 20–30 checks on every mergeManaged hybrid or Postgres FTS plus pgvectorHybrid index per audience behind MCPSimple vector RAG on Vectorize or pgvector
SMBPostgres lexical plus vector, owners per domainRetrieval as a service, ACLs before retrievalAuthorship and permissions become the bottleneckPR-based Postgres pipeline, or Copilot
Mid-marketCurated brain plus separate employee searchFederate sources; buy or buildBuy search, build a smaller canonHybrid plus GraphRAG, or Glean
EnterpriseCurated brain plus procured search; no whole-company crawlEnterprise search control plane, or build for sovereigntyIdentity-aware platform, federated canonsGlean, Copilot or custom multi-agent
Eval setStart at 20–30, grow per domain20–30 is a smoke test only100–300 stratified20–30 is "highly mature"
Biggest gap foundNo written file contractMCP is not authorization; 11 missing controlsEU residency fails for the index layerObsidian friction; no graph
On its own makerxAI Collections priced like OpenAI's; "a component, not the source"OpenAI file search fine for prototypes, keep the corpus outsideDiscloses it is an Anthropic model; memory tool is not governed knowledgeNot addressed

Positions as each model wrote them, shortened. Cost ranges are each model's own planning estimates.

Where they agree

  • All 4
    The architecture is the right shape

    Files as source, a rebuildable index, live data via MCP and human approval all survived every review.

  • All 4
    Upkeep beats technology

    Each model says brains die from neglect, not from bad search, and that tokens and people cost more than infrastructure.

  • All 4
    Long context only for small corpora

    "Just load the files" is right for a micro business and wrong as the plan for thousands of notes.

  • 3 of 4
    Cloudflare residency is not EU-only

    Grok, GPT and Claude independently found Vectorize lacks an EU jurisdiction. Gemini did not check.

  • 3 of 4
    Test Greek, don't trust the card

    bge-m3 lists Greek but publishes no Greek retrieval score; run real Greek, unaccented and Greeklish queries.

Where they disagree

  • Split
    Graphs

    Gemini wants GraphRAG from mid-market. Grok cites a 2026 study where graph builds stayed below BM25 at scale; GPT and Claude say links and tables first.

  • Split
    How big the eval set must be

    Gemini calls 20–30 questions highly mature. Claude calls it statistically unable to detect regressions.

  • Split
    Copying live data

    Three models endorse "never copy". GPT says that is too absolute: allow labelled caches with a TTL for outages and analytics.

  • Split
    Index rebuilds

    Gemini warns full rebuilds get slow and wants incremental indexing; Grok and Claude want full rebuilds in CI against a staging index, then a swap.

  • Split
    Copilot's price

    Gemini and Claude quote $21 (Business); GPT and Grok quote $30 (Enterprise). Both are right for their plan.

FAQ · the questions people ask about agent knowledge bases

Questions, answered

  • Q
    What is an agent-readable second brain?

    The single, governed knowledge base a company's AI agents read from: procedures, rules, decisions, product facts and approved answers. Live data such as bookings and invoices stays in its own systems.

  • Q
    Do small businesses need a vector database?

    Usually not at first. Under about 20 people, markdown files in git that the agent reads directly are enough. Add a hybrid index when a second audience, a few hundred notes or failing tests appear.

  • Q
    Is Obsidian a good choice?

    As an editor for the people who curate, yes. The real source of truth is the markdown in git; non-technical staff need a simpler proposal form, and shared vaults cap at 20 collaborators.

  • Q
    Why hybrid search instead of vectors?

    Keyword search catches product codes, names and Greek inflections that embeddings miss. Anthropic measured a 49% cut in retrieval failures from adding BM25, and 67% with reranking.

  • Q
    Can a Cloudflare stack keep all data in the EU?

    Not fully, as of October 2026. R2, D1 and Durable Objects can be pinned to the EU; Vectorize, AI Search and Workers AI have no documented EU jurisdiction. Use EU-region Postgres for the index if you need EU-only.

  • Q
    Should we just buy Copilot or Glean?

    For employee search across mail and documents at 100+ people, often yes. For customer-facing or rule-bound agents, keep a curated canon you own beside it.

House of 47 read · How to think about it

Invest in the gate, not the index

Embeddings are cheap, owners are scarce

bge-m3 embeddings cost about a cent per million tokens. The scarce input is a person who can say of one page: this is the rule, I own it, it is still true. Fund that person before a second vector store.

Split audiences on day one

The incident that hurts is a public bot quoting a margin or a salary. A separate public folder, index and credential costs almost nothing at five people and is very expensive to retrofit at fifty.

Keep the canon portable

Hosted file search is priced almost identically across OpenAI and xAI, which is what a commodity looks like. Teams with their truth in git re-embed and continue; teams with it in a vendor index migrate again. For a broader view of model choice, see Best AI models of 2026.

Residency is per product

"The bucket is in the EU" is not the whole story: the index, the embedding call, the generation call and the logs each have their own location. Write every processor into the DPA, as the design notes already say to check.

Base case

Plan for a small, typed, git-backed canon, a rebuildable hybrid index per audience, live systems behind read-only MCP, and a growing test set that gates every change. Buy suite search beside it once employees, not agents, are the main askers.

Tail

Stress-test for the swings. A poisoned document that slips past a tired reviewer, a cross-audience leak, or a regulator asking where prompts are processed would each force a redesign. Cheaper million-token context could also push the "just load it" line far higher for clean handbooks.

The brain that lasts is not the one with the best index. It is the one where every page has an owner and every change has a test.

Sources · figures as cited by the four model reports, 14 October 2026

Sources

Anthropic 2024, Contextual Retrieval (35% / 49% / 67% fewer failures), study

arXiv 2607.26497 2026, file agents vs BM25 vs graphs at enterprise scale, paper

arXiv 2607.21274 2026, CUP Greek retrieval benchmark (hybrid 0.673 nDCG@9), paper

Chroma, Context Rot across 18 models, research

Liu et al. TACL 2024, Lost in the Middle, paper

arXiv 2501.01880, long context vs RAG revisited, paper

Microsoft Research, LazyGraphRAG cost and quality, blog

arXiv 2502.11371, RAG vs GraphRAG systematic evaluation, paper

OWASP LLM01 Prompt Injection, Top 10

OWASP LLM08 Vector and Embedding Weaknesses, Top 10

OWASP RAG Security Cheat Sheet, cheat sheet

MCP specification 2026-07-28, OAuth 2.1 authorization, spec

European Commission, AI Act Article 50 transparency FAQ, FAQ

European Commission, GDPR principles, guide

Gartner 2025, over 40% of agentic AI projects cancelled by 2027, press release

Microsoft, Copilot oversharing guidance, blog

Cloudflare, Data Localization Suite compatibility, table

Cloudflare, Vectorize overview and pricing, docs, pricing

Cloudflare, AI Search limits and pricing (billing from 1 Nov 2026), docs

Cloudflare, R2 data location and pricing, location, pricing

Cloudflare, D1 limits, docs; Workers pricing, docs

Cloudflare, Workers AI pricing (bge-m3 $0.012/M), docs

Cloudflare, AI Gateway pricing and logs, docs

Cloudflare, Access pricing, pricing

SQLite, FTS5 diacritics handling, docs

PostgreSQL, text search configurations incl. greek, docs

MTEB, BAAI/bge-m3 model card, card

Obsidian, pricing and Sync collaboration limits, pricing, docs

GitHub, Team plan, pricing

Supabase, Pro plan, pricing

Pinecone, Builder and Standard, pricing

LangSmith and Langfuse, tracing plans, LangSmith, Langfuse

OpenAI, API pricing incl. file search, pricing; business plans, pricing

xAI, Collections pricing, pricing

Anthropic, Claude API pricing, pricing; memory tool, docs

Google, Gemini File Search pricing, pricing

Microsoft, 365 Copilot Business ($21) and Enterprise ($30), business, enterprise

Notion, Business plan ($20), pricing

Glean, Enterprise Flex (no public price), docs; Vendr median contract, marketplace

Changelog · what changed in this brief, newest first

Changelog