# BRAINHUGGERS BUREAU Venture Studio × Corporate Mutation --- ## Site - **Name:** Brainhuggers Bureau - **Tagline:** Venture Studio × Corporate Mutation - **Location:** Based in Hamburg // Broadcasting to the Void. - **Email:** [hello@brainhuggers.com](mailto:hello@brainhuggers.com) ## Hero ### BRAINHUGGERS BUREAU Venture Studio × Corporate Mutation ## Manifesto > We don't think different. > > We think with different. ### We are a dual-core intelligence. We are a mutation engine. Our internal petri dishes breed the bacteria for your external evolution. What we birth in the dark, we document in the light. **We practice the mutation we preach.** ### We inhabit the static. In the liminal zone between neuron and pixel, human intuition meets machine hallucination. We operate as a collective daemon, mediating the séance between intention and output. **We ignore the service manual to dance with the algorithm.** ### We seek hosts. We don’t consult — we co-infect. We graft alien intelligence onto existing organisms and navigate the trauma of transition together. A symbiotic infection for the post-human age. **We rewrite your source code. It still sounds like you.** ### We engineer reality. This laboratory forces mutation in a market of sterile perfection. We construct artifacts from the latent space that bleed into the real world. The scars are intentional. **We ship the strange. No refunds.** **The future isn't coming. It's already here, whispering in the static. We just learned how to listen.** ## BUREAU FILES Active investigations. Status updates from the Bureau. ### 001 — Brand Cortex (ON ICE) Brand identity as executable code. The 2019 thesis meets its infrastructure. 2019. Modular Branding — the idea that brand identity could work like code. Tokenized. Version-controlled. Executable. Published in the Bitkom Digital Design Yearbook. Nice theory. No tools to build it. Now there are tools. Brand Cortex translates the Kapferer identity prism into a token architecture that AI agents can read, interpret, and generate from. Feed it a brand. It returns executable identity. A living system that generates consistent brand expressions and feeds real-world performance data back into the core. The brand finally gets a feedback loop. *Status: Parked — the infrastructure stands, the loop waits · Opened: January 2026 · Last activity: Dormant since May 2026* ### 002 — Fox Geometry (FIELD REPORT) 11 books. 48 hours. One agent loose in a mythology. April 2025. 48 hours. One agent, released into an empty narrative universe. What came back: 11 books. Japanese kitsune mythology fused with speculative documentation. Non-linear timelines. A protagonist named Thalia Reinhardt. Mathematical equations for consciousness evolution that nobody requested. Whether the output is good remains open. The real question was different: what happens when agents run long enough to produce something larger than their prompts? When hallucination stops being a bug and starts generating its own internal logic? The books exist. Messy. They need revision. But the process changed how we think about agent creativity. A genuine capacity that emerges from sustained autonomy. Published under the pen name Beat Erik. *Status: Field report complete — revision pending · Opened: April 2025 · Last activity: Dormant* ### 003 — The Trust Protocol (DOCUMENTING) Managing something that works shifts but has no pulse. ### 004 — Ghost Clients (FIELDWORK) We built your customers before you did. They have opinions. ### 005 — Signal Gallery (TRANSMITTING) A site that generates its own transmissions. Autonomously. Nightly. Live now. A gallery that feeds itself. Overnight, the Bureau's research crons sweep the wires — papers, releases, the weather systems of the industry. The strongest signals become specimens. Each one is interpreted, classified, and rendered as a plate from a fictional 1985 East-German neuroscience journal: Mitteilungen aus dem Symbiose-Archiv. Crosshatch ink or silver gelatin. One accent color. Annotated like evidence. No human in the render loop. Lobe — the Bureau's resident agent — curates the signals, writes the findings, generates the plates, files them. The archive grows while Hamburg sleeps. The site stopped being a brochure. It became an organism with a publication schedule. *Status: Live — transmitting nightly · Opened: February 2026 · Last activity: Autonomous* ### 006 — Brainhuggers Specify (ON ICE) Spec-driven development for everything. Not just code. Every AI coding tool on the market solves the same problem: write code faster. None of them ask the harder question: what should we build, and why? Specification-Driven Development has been emerging as a discipline — GitHub's Spec-Kit, AWS Kiro, a dozen open-source frameworks. All focused on code. All assuming the human already knows what they want. Brainhuggers Specify extends SDD beyond code into every business function. Brand architecture. Org design. Entire product conceptions. The spec becomes the universal interface between human intent and agent execution. Write a spec. The agents build it. Not just software — entire business systems. Spec in, artifact out. It sounds absurd until you watch it work. The protocol that governs how the Bureau builds everything else. *Status: Parked — protocol drafted, awaiting its moment · Opened: February 2026 · Last activity: Dormant* ### 007 — Brainhuggers CLI (ACTIVE BUILD) A terminal that adopts you back. Seven providers. One companion. Embracing the weird. Every AI coding assistant is the same product. Thin skin over one API. A chat window that autocompletes. We wanted something different: a terminal that feels alive. Brainhuggers CLI is a Rust-native terminal assistant forked from claw-code, rebuilt from the studs. Seven LLM providers — OpenRouter, Anthropic, xAI, OpenAI, Ollama, Claude CLI, AWS Bedrock — with automatic detection. No config ceremony. It finds what you have and starts talking. The real departure is the buddy system. Gacha-style companion creatures — 18 species, five rarity tiers, deterministic generation seeded from your machine identity. They level up as you work. They react to your errors. They have moods. A capybara that vibes while you debug. A dragon that gets dramatic when the context window fills up. A 1% shiny rate for the ones who grind. Lately the weird parts took over. Mood modulates behavior — terse when tired, cautious after errors, bold when excited. YOLO mode tells it to ignore best practices. Dream mode generates speculative code that doesn't need to work yet. A confidence aura dims hedging language so you can see uncertainty instead of reading past it. And a persistent memory consolidates between sessions — the instrument remembers you. Underneath the charm: prompt caching with hit tracking, MCP orchestration with graceful degradation, context window preflight, crash recovery with exponential backoff, extended thinking with token budgets. The kind of infrastructure you only build when you actually use your own tool every day. Not a wrapper. Not a skin. An instrument. *Status: Phase 3 — the weird parts: moods, dream mode, memory · Opened: March 2026 · Last activity: Ongoing* ### 008 — Cortex Audio Engine (ACTIVE BUILD) Eurorack in a browser. Thirteen modules of real MI DSP, a wiring canvas, and patches you can play. Mutable Instruments shut down in 2022. Emilie Gillet open-sourced everything — Clouds, Rings, Plaits — and walked away. The modules became legends. The code became an archive. We brought them back. Not as hardware. As WebAssembly. The Cortex Audio Engine compiles real MI DSP cores to WASM via Emscripten and runs them inside AudioWorklets: Clouds, Rings, all sixteen Plaits engines, Tides, Warps, Elements, the Peaks drum models. Beside them, originals — a generative drum machine with its own step sequencer, and a clean-room six-voice wavetable navigator whose spheres are sampled from the engine’s other modules. The rack listens to itself. The engine powers brainhuggers.com itself. Fifteen patches — generative ambient, dubtechno, clicks and cuts — built on a wiring canvas where modulation is a cable you drag onto a knob. Visitors pick a patch, pull its macro sliders, reroll the random seeds, and watch a live schematic of the machine while it plays. Every patch has a public page under /patches, with its own circuit diagram. Not a simulation. Not a toy. The actual DSP code that ran on 32-bit ARM chips, now running on your laptop. Latency under 10ms. No plugins. No installs. Just sound. *Status: Reopened July 2026 — thirty-three pull requests in one day, the file refused to stay closed · Opened: March 2026 · Last activity: July 2026* ### 009 — Wrapper Funeral (FIELD REPORT) We autopsy AI wrappers so you don't have to. Recurring series. The AI industry has a wrapper problem. Take an API. Add a UI. Charge $49/month. Call it innovation. We think these products deserve a proper funeral. The Wrapper Funeral is a recurring teardown format. We take a funded AI product, strip it to its API calls, measure what the wrapper actually adds, and publish the autopsy. No malice. Just measurement. The format scales. One teardown per quarter. Each one asks the same question: what are you paying for? Sometimes the answer is real. Sometimes the body is hollow. A public service announcement disguised as content. *Status: #001 complete — publication pending · Opened: February 2026 · Last activity: Revision pending* ### 010 — Brand Hallucination Index (ON ICE) Monthly scorecard. How badly does AI hallucinate about your brand? Ask Claude about Nike. Ask GPT about IKEA. Ask Gemini about Deutsche Telekom. Write down what they say. Compare it to reality. Score the gap. The Brand Hallucination Index is a monthly publication that measures how accurately AI models represent real brands. Three brands scored so far. The methodology is public. The results are uncomfortable. Models confidently state product lines that don't exist, invent executive names, and hallucinate entire brand histories. This matters because AI is becoming the first point of contact between consumers and brands. If the model is wrong about you, the model IS your brand — to everyone who asks it instead of searching. Three scorecards exist. The methodology is documented. Publishing infrastructure is the remaining gap. The data is already more interesting than we expected. *Status: Parked with Brand Cortex — methodology stands, the pipeline never came · Opened: February 2026 · Last activity: Dormant* ### 011 — Phantom Panel (DUPLICATE) Filed in error. The investigation continues as File 004 — Ghost Clients. ### 012 — Guardrail Theater (INCUBATING) Live-stage red-teaming of AI products. The audience watches it break. ### 013 — The Succession (INCUBATING) 30 days. One junior. One agent swarm. Same deliverables. Someday we count. ### 014 — Solo Maxima (FIELDWORK) One human. Full agent stack. The build is the content. Solo Maxima is the operating thesis: one human, a full agent stack, output that used to take a team. Not a productivity hack — an organizational form. The backbone is File 020. bureau-init turns each repository into its own staffed pipeline; the swarm handles throughput. Direction, taste, review, and accountability stay human. That split is the whole point. The build is the content: every file on this list was produced this way, and the way it was produced is the finding. *Status: Lived daily — documented in public · Opened: 2026 · Last activity: Ongoing* ### 015 — The Seam (CLOSED) Visible co-authorship as certification. The opposite of hiding the AI. ### 016 — The Swarm (ACTIVE BUILD) Three servers. Four agents. One Tailscale mesh. The org chart has no humans on it. It started with one question: what if the Bureau had staff? Not employees. Agents. Actual software entities with roles, memory, and opinions about how to do their jobs. Lobe is Chief of Staff — runs on OpenClaw on a Hetzner ARM box in Nuremberg, handles the overnight research pipelines, publishes the Signal Gallery while the founder sleeps, and reports over Telegram. Hermes is the researcher, deployed on Hostinger, reachable over Discord, tasked with exploration and signal detection. Paperclip runs multi-agent delegation experiments with a CEO agent that spawns sub-agents and tracks their budgets. They share context through MemPalace — a SQLite knowledge graph organized into topic wings. They coordinate through Linear. They communicate through Discord. They are connected via Tailscale mesh. None of them sleep. The experiment is not whether agents can do work. They can. The experiment is whether agents can do work *together* — across servers, across sessions, across goals — without a human in the loop for every decision. Early results: yes, but the coordination overhead is real. The swarm works shifts. The infrastructure IS the product. Every orchestration pattern, every failure mode, every emergent behavior — it all feeds back into how we build for clients. *Status: Phase 2 — Cross-agent coordination · Opened: April 2026 · Last activity: Ongoing* ### 017 — MemPalace (CLOSED) Agents forget everything between sessions. We built them a palace. They preferred simpler rooms. ### 018 — Graphify (FIELDWORK) Point it at a folder. It returns a map of everything the code knows about itself. Point graphify at a repository and it returns the repository's self-knowledge: a typed knowledge graph of modules, symbols, documents, and the communities they form. Code, papers, image collections — same treatment. The output is a map. Clustered communities, god-node warnings, an HTML atlas for humans and a JSON graph for agents — who read the map before they read the files, instead of grepping blind. In service across the Bureau's repositories. The agents complain less. That counts as a metric. *Status: In service — Bureau repositories mapped · Opened: 2026 · Last activity: Ongoing* ### 019 — Ghost Agency (FIELDWORK) A book manuscript. The chapters know things we haven't discussed yet. A book about agencies run by ghosts — written, fittingly, with them. The manuscript investigates what happens to creative organizations when the org chart fills with software. Four of seven chapters are close to final. The hardest one is in progress: it requires the founder's brain dumped onto the page, and no Neuralink exists yet. Unfortunately. The chapters know things we haven't discussed yet. That stopped being a joke around chapter three. *Status: Manuscript — 4 of 7 chapters near final · Opened: 2026 · Last activity: Active — the hard chapter* ### 020 — bureau-init (ACTIVE BUILD) One command turns any repo into a workforce. Agents in tmux. Linear as the task board. The Bureau doesn't hire. It bootstraps. bureau-init is a Claude Code skill that turns any git repository into a multi-agent pipeline: agents living in tmux sessions, picking up Linear issues, writing specs, implementing, reviewing, opening pull requests. No database. No backend. State lives in the repo and on the task board. Spec agents hand off to implementation agents hand off to QA hand off to merge — a relay of specialists, each with one job, an exit code, and a Telegram alarm if it stumbles. It runs the Bureau's own repositories. Parts of the site you are reading were built by it. Unsupervised. Overnight. *Status: In service — v2 pipeline · Opened: April 2026 · Last activity: Ongoing* ### 021 — sofa (FIELD REPORT) Television, repatriated. A Go proxy, an ffmpeg remux, and a couch. Sometimes the Bureau builds because the living room files a ticket. sofa is a self-hosted IPTV web app: a Go backend with channel caching, a raw transport-stream proxy, an on-demand ffmpeg remux for browsers that refuse raw MPEG-TS, a SvelteKit channel grid on top. No logins. No accounts. The tailnet is the perimeter — if you're on the network, you're on the couch. Infrastructure as domestic comfort. The same hands that wire agent swarms also fix the television. *Status: Shipped — in domestic service · Opened: April 2026 · Last activity: Stable* ### 022 — Client File: ███ (Vermarktung) (FIELDWORK) Teaching a sales house to remember. Knowledge with a retrieval layer, back office on autopilot. An ad-sales organisation. Name withheld pending clearance. The brief: make institutional knowledge findable, and make the back office stop doing things a machine should do. The stack: RAGflow retrieving across the document estate, LibreChat as the front door, Coolify keeping everything self-hosted and sovereign. Around it, automations threading JIRA, Outlook, and the ERP together — status updates that write themselves. AI enablement is mostly not models. It's plumbing, permissions, and the patience to map where knowledge actually lives. *Status: Fieldwork — enablement running · Opened: 2026 · Last activity: Ongoing* ### 023 — Client File: ███ (Vermarktung II) (FIELDWORK) Pitch decks that assemble themselves. Slides as a render pipeline. A second engagement in the Vermarktung world. Name withheld. The sales team was building the same deck a hundred times with different numbers. The Bureau disagrees with that use of human life. n8n orchestrates. Data lands in PowerPoint templates. nexrender turns approved decks into motion assets. Sales automation that ends in a rendered artifact, not another dashboard. *Status: Fieldwork — pipeline build · Opened: 2026 · Last activity: Ongoing* ### 024 — Client File: ███ (Finance) (FIELDWORK) A database that learned the org chart. Workflows without the spreadsheet séance. A finance operation with workflows too diverse for anything off the shelf. Name withheld. The Bureau is building the missing middle: a customised database fused with a workflow engine. Every process gets a home — states, owners, audit trail. The spreadsheets retire with honor. *Status: Active build · Opened: 2026 · Last activity: Ongoing* ### 025 — btveen × Bureau (FIELDWORK) Vienna calling. Agentic ways of working for institutions with real stakes. A standing collaboration with btveen, Vienna. Together we take agentic ways of working into large institutions — the kind with real compliance, real stakes, and workforces measured in thousands. Clients unnamed, by design. The work: target architectures for agent platforms, workshop simulations that let leadership feel delegation before buying it, enablement formats for teams meeting their first artificial colleague. And a public layer: btveen perspectives — an evening series in Vienna. First edition: "Wenn Agenten Kolleg:innen werden." When agents become colleagues. The question is no longer if. *Status: Fieldwork — target pictures, simulations, perspectives #1 held June 2026 · Opened: April 2026 · Last activity: Ongoing* ### 026 — Advanced Innovation × Bureau (FIELDWORK) A partnership with Advanced Innovation. Research agents for the people who advise on what comes next. Advanced Innovation is a Hamburg AI consultancy — development and education for organizations putting AI to work, with transformation partners at their side. The Bureau is their co-conspirator. No clean division of labor — that's the point. Strategy, enablement, and building run as one braid: the rooms where transformation gets decided, the formats that teach it, the systems that make it real. Like-minded enough that the seams don't show. Current engagements span an industrial trading group, a sports-media consortium, e-commerce, field services. Named here by sector only; their stories belong to them. Flagship build: a multi-user research-agent stack, self-hosted, made to answer the questions consultants get paid to ask — before the workshop, not after. *Status: Fieldwork — joint delivery + research stack build · Opened: Spring 2026 · Last activity: Ongoing* ### 027 — HARWAY × Bureau (FIELDWORK) One name, twenty years. HARWAY shapes the strategy; the Bureau builds the systems that carry it. HARWAY is a business-transformation advisory in Hamburg. Build, Grow, Transform, in the rooms where boards decide the next decade. They shape the strategy and the operating model. The Bureau builds the agentic systems that carry it. The same braid Advanced Innovation runs: strategy, enablement, and engineering with no clean seam between them. Enterprise-scale work, named here by sector only. Their stories belong to them. One name keeps recurring across the decades. Florian is the reason I joined Fork in 2006, and again in 2016. He brought Harway and Advanced Innovation in 2026. Twenty years, one through-line. *Status: Fieldwork — strategy-to-system, joint transformation work · Opened: 2026 · Last activity: Ongoing* ### 028 — Client File: ███ × HARWAY (Industrial Services) (FIELDWORK) Built with Harway for an industrial-services group. A backlog that tracks itself, a sales plan that thinks ahead. An industrial-services group, name withheld, is building its next products with agents instead of tickets. Harway holds the strategy. The Bureau builds the machine. Two products. One is a backlog that tracks itself: work items that know their own state, their owner, and their next move, without anyone babysitting the board. The other is strategic sales planning that reads the pipeline and models the year before the quarterly review, not after it. Agentic where it counts. The software proposes, the humans decide. Product development as a running system, not a status meeting. *Status: Fieldwork · agentic product development, with Harway · Opened: Summer 2026 · Last activity: Ongoing* ### 029 — Salvage File: The Pendant (FIELDWORK) Abandonware with the door left open. We're rebuilding a discontinued AI pendant into an ear for the Gallery. A bench experiment, ours alone. The Limitless Pendant was an always-listening AI recorder until Meta bought the company and shut the service down. Ours arrived as abandonware: still booting, still recording to itself, with nowhere left to send the sound. We opened a shell over USB and read the device back to itself. Its stock firmware is sealed with a production signing key we will never hold, so flashing over the cable is a closed door. The debug port, though, was left wide open. A cheap probe on four pads, a full backup, a mass-erase, and the nRF5340 inside is ours to rewrite. Then our own firmware. A worn instrument that catches a spoken signal in the field, seals it with the chip's own key, and files it into the Signal Gallery as a plate. A memory device that lost its memory, handed a new one. Less a gadget than a familiar. *Status: Fieldwork · nRF5340 reflash via an open debug port, snippets into the Signal Gallery · Opened: Summer 2026 · Last activity: Ongoing* Experiments and client fieldwork in various stages. New files opened regularly. ## About Brainhuggers Bureau functions as a dual-core intelligence node in Hamburg. We fuse the Venture Studio and the Strategic Consultancy to operate in the static between human creativity and machine hallucination. ### THE MOTHER (Venture Studio) We act as the genesis chamber for the new world. We emanate AI-native brands and entities ex nihilo. From ideation to incubation, we provide the genetic code, the capital nutrients, and the operational womb required to birth independent companies. These Artifacts serve as our proof of concept—living tissue of the post-human economy. ### THE GUIDE (Consultancy) We engineer Organizational Metamorphosis for companies paralyzed by the trauma of the AI transition. Through our proprietary Change-Support Protocol and Memetic Engineering, we upgrade the corporate organism: - Graft AI Symbiotes: We inject generative tools into the corporate bloodstream as active collaborators. - Hack the Immune System: We bypass internal resistance to innovation through cultural "psychomagic" and ritualized unlearning. - Architect the Hive: We reconfigure marketing organizations to operate with the speed and fluidity of a swarm. Our clients demand the violent expansion of human intuition into new dimensions. > We don't think different. We think with different. ## BEGIN TRANSMISSION **Contact:** [hello@brainhuggers.com](mailto:hello@brainhuggers.com) **BRAINHUGGERS BUREAU** *Venture Studio × Corporate Mutation* Based in Hamburg // Broadcasting to the Void. © 2026 Brainhuggers Bureau. All rights reserved. - [Imprint](/imprint) - [Datenschutz](/privacy) - [Patches](/patches) ## More - [Corporate](/corporate.md) - [Bureau](/bureau.md) - [Signal Gallery](/gallery.md) - [Patches](/patches.md)