Trust center Public posture Limits included
Trust is an interface requirement
Know what this site does, what it keeps, and where its authority stops.
This center documents the current public posture for privacy, security, accessibility, evidence, and engineering decision boundaries. It describes the implementation honestly; it is not a certification.
Current posture
Four trust commitments, with their limits
Each statement is scoped to what this public repository and its application code can support.
Privacy
Technical inputs remain browser-local
The workbench is built for public-safe, preliminary structuring without an application data pipeline.
- Cookies
- None in site code
- The first-party application code does not set or read cookies.
- Usage analytics
- Not installed
- The public application does not include an analytics or behavioral-tracking integration.
- Technical inputs
- Browser-local
- Interactive tool values are processed in current-page browser memory. There is no application endpoint or database receiving those values.
- Persistent input storage
- Not used
- The tools do not use localStorage, sessionStorage, IndexedDB, or a site-side input store.
- Email drafts
- Manual only
- A user can intentionally open a prepared draft in their own email client. The site does not send it automatically.
- External services
- Separate policies
- Hosting infrastructure and destinations reached through external links may process normal request metadata under their own policies.
Security
A narrow static surface, not a security guarantee
Static delivery removes application accounts and a site-side input database, but it does not remove infrastructure, browser, dependency, DNS, TLS, build-chain, or third-party risk.
Architecture
Static public files
The deployed application is a statically generated site. It does not provide accounts, login, application authentication, or a server-side input store.
Inspectability
Public source repository
Page code, browser tools, public data, and deployment configuration can be inspected in the public repository.
Honest limit
No audit claim
This page does not claim a penetration test, independent security audit, security certification, uninterrupted availability, or absence of vulnerabilities.
Responsible reporting
Share the minimum safe reproduction.
Start through the contact page. Use a public repository issue only for non-sensitive, reproducible problems. Do not publish credentials, personal data, confidential material, or live exploit details; request a private contact route first.
Accessibility
Designed for access; open to correction
The site aims to support keyboard, screen-reader, zoom, motion-sensitive, and small-screen use as part of normal product quality.
- Skip-to-content navigation and semantic page landmarks
- Keyboard-operable primary navigation, mobile navigation, and site search
- Visible focus treatment for interactive controls
- Labels and state cues that do not depend on color alone
- Responsive layouts and reduced-motion behavior
- Concise text alternatives or hidden treatment for decorative visuals
Claim and source governance
Evidence should travel with the claim
The public trust model separates sourced facts, interpretation, limitations, and material that is still held for review.
Source context
A new factual claim needs public source context. Analysis or interpretation should be identified as such.
Claim registry
Proof metrics and identity facts are routed through shared data, with source status, limitations, and review context.
Visible boundaries
Monitoring signals, model confidence, completeness labels, and evidence-burden labels are kept separate from approval.
Correction path
The public claim ledger exposes source context and a route for suggesting corrections.
Decision boundaries
Decision support is not decision authority
Structured context can improve review quality. It cannot substitute for material facts, inspection, acceptance criteria, or accountable expert judgment.
Site-wide disclaimer
Preliminary decision-support only. Final feasibility depends on base material, geometry, service conditions, inspection requirements, and expert review.
Not valid for engineering approvalNot valid for certificationNot valid for final releaseNot valid for safety-critical acceptanceNot valid for quality guarantee
Context quality
Completeness is not feasibility.
A more complete brief means the review context is better structured. It does not mean the proposed route is feasible.
Evidence planning
Evidence burden is not approval.
An evidence-burden label helps plan review and inspection. It is not acceptance, release, or a quality guarantee.
Commercial boundary
Exafuse owns commercial review.
Services, RFQs, production capability, delivery claims, company case studies, and quality pages belong to Exafuse.
Contact paths
Route the issue without crossing a boundary
Choose a public-safe route, and keep confidential or safety-critical material out of this personal site.
Security / privacy / accessibility / correction Contact route selector Start here to choose LinkedIn, GitHub, or Exafuse based on the issue. Choose a route Professional discussion LinkedIn Use for public professional context, technical discussion, and non-confidential collaboration questions. Open LinkedIn Industrial RFQ / services Exafuse contact Use the company route for commercial review, RFQs, services, production, delivery, and company-owned claims. Open Exafuse
Identity path
Continue through the public identity layer
Industrial AI for Decisions You Can Verify The operating thesis behind the public lab. Open → Manish Sharma - Industrial AI & Decision Systems The canonical public identity page. Open → LMD / DED Domain Hub The established public proof domain. Open → Full Site Map Every page and public machine-readable asset in one linked map. Open →