docs(blog): retire local-first present-tense claims across posts#234
Merged
operatoruplift merged 1 commit intomasterfrom Apr 28, 2026
Merged
Conversation
Eight surgical edits to app/blog/[id]/page.tsx that align the blog narrative with what the web app actually ships today vs what is roadmap-only. The web app is Vercel-hosted and routes prompts through whichever AI provider the user picks per turn. The local-first product is a desktop+Ollama path on the roadmap. Edits: - balaji-pivot-advice (line 16): "a local-first AI operator" -> "an AI operator with approval-gated tool calls, ed25519-signed receipts per action, and a roadmap toward a desktop+Ollama build for full local inference" - local-first-threat-model (line 70): rewrite the "raw content stays on your machine" paragraph to describe the real provider hop, the localStorage cache for chat sessions and memory, and the desktop+ Ollama roadmap as the path that removes the hop. - local-first-threat-model (line 73): rewrite the "API keys live encrypted in your browser via AES-256-GCM" paragraph to describe the actual state: the passphrase setup is wired but secureStore/ secureRetrieve are not yet called from chat sessions or memory. Disclosure matches the in-product copy on Settings -> Security. - wedge-lawyer-accountant-therapist (line 89): rewrite the "encrypted local memory ... never touch a third-party cloud" sentence to describe the per-turn provider routing, the approval gate, and the on-chain audit trail. Mention the desktop+Ollama path as the roadmap option for full on-device inference. - why-i-built-an-ai-os (line 108): "a local-first AI operating system" -> "an AI operating layer with approval-gated tool access and a local-first roadmap" (preserves the past-tense framing of what was shown to Balaji without overstating shipped state). - why-i-built-an-ai-os (line 113): rewrite the "It runs on your device. Your agents, your memory, your context, encrypted, local, yours" paragraph to describe what the web app does ship today (approval gate, signed receipts) and what the desktop build adds (local inference via Ollama). Preserves the OS-layer / landlord vs tenant framing. - what-93-percent-retention-looks-like (line 126): rewrite the "data lives on your device" / "without phoning home to a server" paragraph around the trust loop (approval + receipt) instead. More accurate, no false present-tense local claim. - generic catch-all post body (line 169): "the local-first AI agent platform" -> "the approval-gated AI agent platform with a local- first desktop build on the roadmap" What stays the same: - The blog's overall position (we are building toward a local-first end state) is preserved everywhere it appears as roadmap. - The OS layer / landlord-not-tenant framing is preserved. - All structural narrative beats remain intact. Verification: - pnpm check (4/4: copy-check, capability-check, trust-gate, fabrication-rot-check) - pnpm exec tsc --noEmit (clean)
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
3 tasks
operatoruplift
added a commit
that referenced
this pull request
Apr 28, 2026
…th (#235) The hero animation rendered "vault sealed memory encrypted" at the end of the COMPLETE phase. The web app does not encrypt chat sessions or memory at rest (encrypt/decrypt callers from secureStore are absent), so "memory encrypted" was a marketing claim drawn straight into the canvas pixel buffer. Replace with "session closed receipt anchored" which describes what actually happens at the end of an action sequence: - session ends cleanly on sign-out / explicit close - the action's ed25519-signed receipt is anchored into the Merkle root that gets published to Solana via the audit-trail Anchor program This continues the honesty positioning sweep landed across the homepage, dashboard, and blog in PRs #227-#234. Verification: - pnpm check (4/4: copy-check, capability-check, trust-gate, fabrication-rot-check) - pnpm exec tsc --noEmit (clean)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Why
What stays
Test plan