BUY $SHRED$SHREDDocsOpen appApp
A private paper document enters a carbon privacy compiler while coral identity fragments stay local and sparse lime context emerges
AI PRIVACY INFRASTRUCTURE LOCAL-FIRST
$SHRED PRICELOADING…MARKET CAPLOADING…PONS LIVE MARKET ↗

YOUR DOC.
YOUR DATA.
YOUR CONTROL.

SHREDDAR compiles sensitive documents into the minimum context AI needs before anything leaves your boundary. Share the task. Keep the source.

TRY A LOCAL PRIVACY RUN ↗TRY THE SAMPLE BELOW ↓
YOUR AI NEEDS CONTEXT. NOT IDENTITY.Browser-local appNo raw telemetry
PRIVACY BOUNDARY / ILLUSTRATED
IDENTITYHELD LOCAL
MEANINGREVIEW BEFORE RELEASE
MINIMUM USEFUL CONTEXT →
PRIVATE WORLDSHREDDAR EDGEAI WORLD ↓
INPUTPRIVATE DATA
→
PRIVACY COMPILER
→
OUTPUTMINIMUM USEFUL CONTEXT
START WITH ONE USEFUL OUTPUT

Less disclosure.
A clearer next step.

The browser alpha transforms text locally. Real model execution belongs behind a separately configured, authenticated gateway.

01 / DOCUMENT REVIEW

Protect your first document.

Open a PDF, DOCX or text file. Set a purpose, inspect the protected text, and verify a metadata-only receipt.

OPEN LOCAL WORKBENCH ↗No wallet · No upload · No external model02 / YOUR INFRASTRUCTURE

Put the boundary beside your data.

Self-host the gateway before connecting a client or agent. Review authentication, provider routing and retention.

READ THE QUICKSTART →Operator setup required · Not a hosted upload service
THE SHREDDAR MACHINE

Identity stays behind the line.

Transform the source, compartmentalize useful meaning, and keep reconstruction within your trusted environment.

Abstract paper data entering a carbon privacy compiler and emerging as protected lime context bandsCONTROLLED DESTRUCTION / 01
COMPILERemove detail. Preserve task meaning.
One abstract document split into three isolated semantic processing routesSEMANTIC ROUTING / 02
SHARDOne task. Bounded routes.
Abstract identity mappings held inside a local carbon vault while protected context travels outwardTRUSTED BOUNDARY / 03
STITCHIdentity mappings stay local.
01

THE EXPOSURE BOUNDARY

Cloud intelligence shouldn't require unnecessary disclosure.

Every model, agent and tool call creates a decision: what actually needs to leave your environment? SHREDDAR makes that decision inspectable and enforceable.

TRUSTED ENVIRONMENTContractsCustomer recordsSource codeWallet operations
RAWSHREDDARPOLICY / VAULT / ROUTEPROTECTED
EXTERNAL INTELLIGENCECloud modelsAI agentsMCP servicesExternal APIs

SEE EXACTLY WHAT YOUR AI SEES. Context crosses the line. Identity does not have to.

02

THE PIPELINE

Controlled destruction.
Useful intelligence.

01

DETECT

Locate identity, secrets, exact values and linkable facts.

02

SHRED

Remove information the requested task does not require.

03

ABSTRACT

Preserve meaning while reducing identifying specificity.

04

SHARD

Create independently routable semantic compartments.

05

INFER

Send only policy-approved protected representations.

06

STITCH

Reconstruct authorized references inside the local boundary.

03

LIVE PLAYGROUND

Inspect the exact context
allowed to leave.

The preview below runs in browser memory. Use synthetic or authorized content—never real credentials. Nothing is sent to our backend.

RUNS IN THIS TAB · ORIGINAL NEVER SENT
01 / ORIGINAL
02 / PROTECTED
Scanning locally…
0 ENTITIES−0% CONTEXT
03 / AI RESPONSE
Waiting for protected context…
MODEL SEES PLACEHOLDERSNO VAULT ACCESS
04 / RECONSTRUCTED
Waiting for authorized reconstruction…
LOCAL STITCH0 MAPPINGS USED
04

THE PRIVACY COMPILER

We don't simply hide names.
We determine what the task needs.

Detection is a component. The product is the decision over relevance, precision, linkability and the minimum useful representation.

ORIGINAL FACTCLASSCOMPILER ACTIONEXTERNAL REPRESENTATION
John SmithPERSONPSEUDONYMIZE<PERSON_01>
Rijksmuseum TwentheORGANISATIONCONCEPTUALIZEregional cultural museum
€82,471FINANCEGENERALIZE€80K–€90K
mira@company.exampleEMAILREMOVE—
API credentialSECRETBLOCK<API_SECRET_PRESENT>
INPUTDOCUMENT + TASK + POLICY→OUTPUTMINIMUM VIABLE CONTEXT
05

SEMANTIC COMPARTMENTS

Split by meaning.
Route by risk.

SHARD separates a task into coherent compartments. Each route receives only its required facts; synthesis happens locally.

PRECISE CLAIM Exposure reduction, not cryptographic secrecy. Correlation remains possible when providers or fragments overlap.

DOCUMENTACQUISITION AGREEMENT
AOBLIGATIONS→MODEL A
BFINANCIAL TERMS→MODEL B
CTERMINATION→LOCAL
LOCAL SYNTHESIS ✓
06

PRIVACY RECEIPTS

Evidence of the operation.
Not a copy of the data.

Every protected operation records policy, provider, transformations, exposure reduction and a verifiable hash—without recording the prompt or mapping.

  • Operational observability
  • Procurement evidence
  • Debugging without raw content
  • Optional onchain hash payload
S/PRIVACY RECEIPT
ILLUSTRATIVE SAMPLE

EXAMPLE ONLY — NOT VERIFIED EVIDENCEAll counts, token totals, risk labels, identifiers and hash fragments below are invented to explain the receipt format. No user document was processed and no provider was called.

REQUESTexample_not_a_request
ProviderExample provider (not called)PolicyExample policyOriginal context18,421 tokensExternal context9,713 tokens
CONTEXT REDUCED47.3%
Sensitive entities42
Pseudonymized21
Generalized08
Conceptualized06
Removed07
Credentials blocked00
RISKHIGH→LOW
RECEIPT HASHsha256:EXAMPLE_NOT_A_HASH

Real receipts contain transformation metadata only—not prompts, identities or Vault mappings. Generate and verify your own receipt in the local Workbench.

07

OPENAI-COMPATIBLE GATEWAY

Change one endpoint.
Protect every request.

Self-host SHREDDAR inside your boundary, point an OpenAI-compatible client at it, and inspect a Privacy Receipt on every response.

INTEGRATE SHREDDAR →API REFERENCE →
terminal
# run inside your trusted boundary
export OPENAI_BASE_URL=http://127.0.0.1:8787/v1

# your existing client
client.chat.completions.create(
  model="configured-upstream",
  messages=[...],
  shreddar_task="Review obligations"
)
RESPONSESHREDDAR RISK: LOW
08

AGENT-NATIVE INTEGRATIONS

Keep your agent.
Add the boundary.

Use the tested MCP server and one portable privacy skill with OpenClaw, Hermes Agent, OpenAI Codex, Claude Code, or another compatible host.

MODELprompt / response→
SHREDDARinspect / transform→
AGENTplan / invoke→
SHREDDARauthorize / minimize→
MCP TOOLbounded context
01 / SINGLE USER

Review locally.

Paste or open authorized text, define the task, and inspect the exact protected output before using it elsewhere.

02 / PROJECT

Route centrally.

Run SHREDDAR inside your infrastructure and point existing OpenAI-compatible clients at one controlled endpoint.

03 / AI + AGENT

Install the guard.

Expose only five agent-safe tools, keep reconstruction outside model context, and attach a Privacy Receipt to every completed operation.

MCP
ANY MCP HOSTFIVE SAFE TOOLS
READY
OC
OPENCLAWMCP + AGENT SKILL
READY
HE
HERMES AGENTMCP + AGENT SKILL
SETUP READY
CX
OPENAI CODEXMCP + REPO SKILL
SETUP READY
CL
CLAUDE CODEMCP + PROJECT SKILL
SETUP READY
ACP
VIRTUALS ACPSERVICE BLUEPRINT
HELD
COPY OR DOWNLOAD YOUR SETUP →COMPARE EVERY INTEGRATION PATH →
09

PRODUCT STRATEGY

Less context in.
Less exposure out.

SHREDDAR is being built as a fast privacy layer before expensive or external intelligence: prepare locally, divide deliberately, and pay only for useful output.

01 / COMPACT LOCAL MODEL

Private before the cloud call.

A compact local model is the planned fast path for conceptualizing ordinary text and documents inside the trusted boundary. That is useful for legal review, e-discovery, diligence, investigations, and other document-heavy work.

DETERMINISTIC ALPHA LIVE / LOCAL MODEL PLANNED
02 / DIVIDE + CONQUER

No provider needs the whole source.

Policy can turn a source into non-overlapping conceptual packages and route each compartment to a different model. Local synthesis recombines results so no single configured provider is intended to receive the full document.

EXPOSURE REDUCTION / NOT CRYPTOGRAPHIC SECRECY
03 / LEAN CONTEXT

Smaller prompts. Lower model cost.

Conceptualization removes irrelevant wording and identifying precision while preserving task meaning. A shorter protected representation can reduce input tokens, latency, and model cost; each receipt will measure the actual reduction.

SAVINGS VARY BY DOCUMENT, POLICY + TASK
US + EU ALIGNMENT

Built to support current rules. Configured for each line of work.

SHREDDAR maps to data minimization, privacy by design, traceability, and risk-management principles in current US guidance and EU law. It is designed to help teams build compliant workflows—not to certify one automatically. Jurisdiction, sector, lawful basis, provider terms, retention, human oversight, and use-case risk still require deployment-specific review.

EU GDPR ↗EU AI ACT ↗US NIST AI RMF ↗PRECISE PRIVACY BOUNDARY →
10 / SECURITY

Fail closed.
Keep mappings separate.

No secrets in prompts. No arbitrary Vault lookup. No raw-content telemetry. Returned output is inspected before local reconstruction.

READ THE THREAT MODEL →
11 / PAY PER USE

Privacy infrastructure
agents can buy per request.

Strict PAYG only: every completed operation is metered by input-context tokens, so larger requests cost more. Pricing and x402 settlement stay held until the rate and route are verified.

EXPLORE PRODUCT + TOKEN UTILITY →

PRIVATE INFORMATION STAYS PRIVATE.
USEFUL MEANING CAN TRAVEL.

SHRED YOUR
FIRST REQUEST.

OPEN SHREDDAR APP ↗SEE HOW IT WORKS →