feat: content restructure phase 1 — Platforms, System, Agents, real Examples#34
Merged
Merged
Conversation
…real domains Phase 1 of the content restructure. The site was missing the pages a serious evaluator actually needs: what is this codebase made of, where does it really run today, and what can an AI agent do with it. All content is grounded in the traverse repo itself (crate manifests, release notes, docs/swift-runtime-candidate-screen.md, package READMEs) — nothing here is aspirational copy. New pages: - /system — the 9-crate workspace grouped by role (runtime core, interfaces, embedding/native bridges, example domain), plus the 73-spec governance layer as an explainer with a link to the real registry on GitHub rather than a browsable spec index. - /platforms — an honest target x status matrix. Native, browser (core + Web embedder SDK), and Rust app embedding are shipped. Swift/iOS is explicitly blocked (no candidate WASM engine exposes the required resource-control APIs, per the July 19 candidate screen), with the wasmi feasibility proof noted as a real but separate workaround. Kotlin and .NET embedders have working bridge code but no public release. Edge is planned; cloud placement is an explicit v0.1 non-goal, not a "not yet" gap. Also flags that "AI pipeline" as a placement target is marketing language ahead of the real runtime target matrix. - /agents — MCP as a first-class pillar instead of a buried guide. The real nine-command tool list, the discover/inspect/execute/trace flow, and the honest stdio-only limitation (no library API yet). - /examples — replaces the single fictional pricing.toml with the six real example domains (expedition, hello-world, doc-approval, meeting-notes, traverse-starter, swift-wasmi-proof), featuring hello-world's actual three-command run path. Nav: Platforms and Agents added as top-level items; The System added to the Product dropdown; Examples' dropdown description updated. Solutions/Use-cases are intentionally untouched here — deletion is phase 2, agreed separately, since those pages are still linked from elsewhere until this lands. components.css: adds badge-danger and badge-subtle, completing the four-state status vocabulary (Shipped/In Progress/Blocked/Planned) used across the new pages and reusable on the roadmap page. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
5 tasks
enricopiovesan
added a commit
that referenced
this pull request
Jul 20, 2026
…s, fix AI-pipeline overclaim sitewide (#35) Phase 2 of the content restructure (follows #34). Delete /solutions/* (fintech, saas, ecommerce, enterprise, ai-companies) and /use-cases — five generic industry-vertical marketing pages for a solo-maintained 0.x project, replaced by the honest /platforms page from phase 1. Every inbound link (Nav, Footer, two blog CTAs, a quickstart card) repointed to /platforms or /system instead. Docs sidebar gains two sections that reuse existing content instead of forking a duplicate copy: "Governance Model" deep-links to the Spec-Driven Development section already on /docs/concepts (now anchored), and "Embed Traverse" deep-links to the five platform rows on /platforms (now anchored per-platform: rust-embedding, browser, swift-ios, kotlin-android, dotnet-winui). /roadmap gets the fourth status this project actually has: a Shipped / Next / Blocked / Exploring legend under the hero, a new "Blocked" phase between "Shipping now" and "Next" surfacing the Swift/iOS certification blocker explicitly (previously invisible on this page), and a fifth principle summarizing the ten explicit v1.0 milestone gates with a link to the source doc. Also fixes a stale "9 governing specs" claim to the real 73. Sitewide fix: "browser, edge, cloud, and AI pipelines" is repeated across the homepage hero/features/stats, about, two compare pages, and five /questions pages as if all four were real, current placement targets. They aren't — native and browser are shipped, edge is planned-not-started, cloud is an explicit v0.1 non-goal, and "AI pipeline" isn't a placement target at all (it's how an agent already reaches the native or browser target over MCP). Corrected every present-tense capability claim to match /platforms; left blog posts alone as dated artifacts and left compare-page problem-framing narratives alone since those describe motivating scenarios, not shipped capability. Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
3 tasks
enricopiovesan
added a commit
that referenced
this pull request
Jul 21, 2026
…est examples (#36) Follow-up to the content restructure (#34, #35), which only touched one overclaiming table row per compare page. A full pass turned up more: Fabricated code samples: all three pages had "integration" code importing a package that has never existed (@traverse-framework/runtime) with an invented TraverseRuntime.init()/.execute() API. A developer copying any of these would hit a wall immediately. - vs-function-calling: the "using them together via MCP" section invented a hand-rolled MCP tool wrapper. The real story is simpler and better — traverse-mcp -- stdio already exposes every governed capability as an MCP tool, no wrapper code needed. Replaced with the real invocation and a link to /agents.html. Also fixed an example contract's placement array, which listed ai-pipeline as a valid value. - vs-microservices: replaced the fake Rust sample with the real traverse-embedder crate API (BundleEmbedder::init, .submit, .subscribe). - vs-serverless: the sample was a Cloudflare Worker, i.e. edge — a placement Traverse hasn't shipped. Reframed as a Rust AWS Lambda handler using the same real traverse-embedder crate, which sidesteps the unverified edge-WASM question entirely by being a real, ordinary Rust process. Malformed HTML: all three pages' CTA bands had a stray </section> landing mid-block (evidently copy-pasted), closing the section early and leaving the button row and final closing tags orphaned outside it. Browsers silently recover from this, but it's invalid structure that happened to look fine by accident. Fixed the tag order in all three so the section actually closes where the markup says it does. Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
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
First phase of a full content restructure. The site was missing the pages a serious evaluator needs: what the codebase is actually made of, where it really runs today, and what an AI agent can do with it. Everything here is grounded in the traverse repo itself — crate manifests, release notes,
docs/swift-runtime-candidate-screen.md, package READMEs — not aspirational copy.New pages
/system— the 9-crate workspace grouped by role, plus the 73-spec governance layer as an explainer that links to the real registry on GitHub (no on-site spec browser, per discussion)./platforms— an honest target × status matrix using four real states (Shipped / In Progress / Blocked / Planned):/agents— MCP promoted from a buried guide to a first-class pillar: the real 9-command tool list, the discover→inspect→execute→trace flow, and the honest stdio-only limitation (no library API yet, issue #310)./examples— replaces the single fictionalpricing.tomlwith the six real example domains (expedition, hello-world, doc-approval, meeting-notes, traverse-starter, swift-wasmi-proof), featuring hello-world's actual verified 3-command run path.Nav
Platforms and Agents added as top-level items; The System added to the Product dropdown; Examples' description updated. Solutions/Use-cases intentionally untouched — their removal is phase 2 (discussed and agreed separately), since other pages still link to them until this phase lands.
components.css
Adds
badge-dangerandbadge-subtle, completing the four-state status vocabulary used across the new pages (and reusable on/roadmap, which already had 3 of the 4).Test plan
npm run build— 93 pages, clean/examples/docs/getting-started.html)🤖 Generated with Claude Code