Add Go SDK examples

[IlyaasK]

Summary

• Add Go SDK examples alongside TypeScript and Python examples across the core Kernel SDK docs: browsers, auth, proxies, browser pools, projects, introduction pages, and the API Keys guide added by PR385.
• Add Go examples for first-party app invocation API pages and the 1Password managed-auth integration where the Go SDK surface is verified.
• Leave app authoring and third-party integration pages TS/Python-only where Go support isn't documented or verified.
• Document Go example conventions in .docs/code-example-guide.md and replace stale README snippet-generation guidance with current inline-MDX ownership notes.
• Clarify that full examples must be runnable, focused snippets can be context-dependent, fake credential-shaped values are allowed only for auth-flow examples, and Go validation must target the minimum released SDK version that supports the documented API.
• Add the API Keys guide and CLI reference nav entries from PR385, with Go examples for create, list/get, update, and delete.

Validation

• Extracted the 3 new API-key Go snippets into a temp module pinned to github.com/kernel/kernel-go-sdk v0.58.0, ran gofmt, go mod tidy, and go test ./....
• Confirmed the API-key Go snippets in info/api-keys.mdx match gofmt output.
• Ran git diff --check origin/main...HEAD.
• Confirmed no changed file crosses the 1k-line review threshold.
• Ran Mintlify preview with npx -p node@22 -p mint mint dev --no-open; checked /info/api-keys and /reference/cli/api-keys returned HTTP 200.
• Previous PR validation: extracted the existing Go code fences from MDX docs plus .docs/code-example-guide.md, formatted them with gofmt, ran go mod tidy in the temp module, and compiled them against the local Go SDK checkout with go test ./....
• Previous PR validation: Mintlify preview checks for /auth/hosted-ui and /auth/programmatic, and autoreview reported no accepted/actionable findings.
• Thermo-nuclear quality review found guidance drift around runnable examples, contextual snippets, fake auth credentials, and stale SDK checkout validation; those issues were fixed in .docs/code-example-guide.md and README.md.

[cursor]

Risk assessment: Low risk.

The diff is confined to .mdx documentation pages and adds Go SDK examples/prose across existing docs. There are no runtime code, configuration, dependency, infrastructure, auth/permission, or schema changes, so the blast radius is limited to documentation correctness and reader-facing examples. I didn’t find a CODEOWNERS file or codeowner review request in the available PR metadata, and there were no existing approvals recorded, so I’m approving.

  

_{Sent by Cursor Automation: Assign PR reviewers}

[mintlify]

Preview deployment for your docs. Learn more about Mintlify Previews.

Project	Status	Preview	Updated (UTC)
Kernel	🟢 Ready	View Preview	May 29, 2026, 3:47 PM

💡 Tip: Enable Workflows to automatically generate PRs for you.

[cursor]

Risk assessment: Low risk.

The synchronized diff is confined to existing .mdx documentation pages and adds Go SDK examples/prose. It does not modify runtime code, configuration, dependencies, infrastructure, auth/permission enforcement, schemas, or prompt/instruction files used by a model. The primary risk is documentation/example correctness across a broad reader-facing docs surface.

I found no CODEOWNERS file and no pending review requests in the available PR metadata. The PR already has an approval recorded, so I’m not adding another approval; the new changes do not increase risk enough to revoke it.

  

_{Sent by Cursor Automation: Assign PR reviewers}

[cursor]

Risk assessment: Low risk.

The current diff is still confined to documentation/contributor guide files (.md, .mdx) and adds or adjusts Go SDK examples and prose. The latest synchronized update expands .docs/code-example-guide.md, updates README.md, and tweaks Go snippets in auth docs, but it does not modify runtime code, site configuration, dependencies, infrastructure, schemas, auth/permission enforcement, or active model/system prompt files.

I found no CODEOWNERS file and no pending review requests in the PR metadata. The PR already has an approval recorded, so I’m not adding another approval; the update does not increase the risk enough to dismiss the existing approval.

  

_{Sent by Cursor Automation: Assign PR reviewers}

[cursor]

Risk assessment: Low risk.

I assessed the current diff directly. The changes are limited to existing documentation/contributor-guide files (.md/.mdx) and add or adjust Go SDK examples and related prose. There are no runtime code, site config, dependency, workflow, infrastructure, schema, auth/permission enforcement, or active model/system prompt changes in the diff.

The broad docs surface and public example correctness risk keep this at Low rather than Very Low, but the blast radius is still limited to documentation. I found no CODEOWNERS file and no pending review requests in the PR metadata. The PR already has an approval recorded, so I’m not adding another approval; this update does not increase risk enough to dismiss the existing approval.

  

_{Sent by Cursor Automation: Assign PR reviewers}

[cursor]

Risk assessment: Low risk.

I assessed the current diff directly. The changes are limited to documentation and contributor guidance files (.md/.mdx), primarily adding Go SDK examples and updating example-authoring guidance. There are no runtime code, site config, dependency, workflow, infrastructure, schema, auth/permission enforcement, or active model/system prompt changes in the diff.

The public docs surface is broad, so example correctness keeps this at Low rather than Very Low, but the blast radius is still limited to documentation. I found no CODEOWNERS file in the repo, and the PR already has an approved review decision, so I’m not adding another approval or dismissing the existing one.

  

_{Sent by Cursor Automation: Assign PR reviewers}

[firetiger-agent]

Monitoring Plan: Go SDK examples added to public docs

What this PR does: Adds Go as a third supported language in Kernel's documentation, giving Go developers working code examples alongside the existing TypeScript/JavaScript and Python snippets.

Intended effect:

• Docs site rendering: No pre-deploy baseline (docs-only change); confirmed if Go code blocks display correctly on the live docs site after Mintlify deploys the merged content.
• Broken-links CI: baseline — must pass; confirmed if .github/workflows/broken-links.yml passes on the merged commit.

Risks:

• Malformed MDX — bad <CodeGroup> tags or unclosed fences could break page rendering; alert if any doc page returns a Mintlify build error or renders blank code blocks post-deploy.
• Invalid Go import paths — github.com/kernel/kernel-go-sdk must be the published module path; alert if the broken-links CI flags import URLs or if user reports of copy-paste failures emerge.
• Stale README reference — removed <OpenAPICodeGroup> generator docs may confuse contributors if the tool is still invoked elsewhere; manual checkpoint: verify no generator CI step references the deleted workflow description.

Status updates will be posted automatically on this PR as monitoring progresses.

View monitor

[hiroTamada]

looks good. But I def need 👁️ from @dprevoznik

[cursor]

Risk assessment: Low risk.

I assessed the current synchronized diff directly. The changes are confined to documentation/contributor guidance files (.md/.mdx) plus docs.json navigation entries for newly added docs pages. There are no runtime code, dependency, workflow, infrastructure, schema, auth/permission enforcement, or active model/system prompt changes in the diff.

The public docs surface is broad and includes many Go SDK examples, so example correctness and navigation accuracy keep this at Low rather than Very Low. I found no CODEOWNERS file in the repo. The PR already has an approved review decision, so I’m not adding another approval, and the update does not increase risk enough to dismiss the existing approval.

  

_{Sent by Cursor Automation: Assign PR reviewers}

[cursor]

Risk assessment: Low risk.

I assessed the current diff directly. The changes are confined to documentation and contributor guidance files (.md/.mdx), primarily adding or updating Go SDK examples and related prose. There are no runtime code, dependency, workflow, infrastructure, schema, auth/permission enforcement, or active model/system prompt changes in the diff.

The surface area is broad across public docs, so documentation/example correctness keeps this at Low rather than Very Low, but the blast radius remains limited to docs. I found no CODEOWNERS file in the repo. The PR already has an approved review decision, so I’m not adding another approval, and this update does not increase risk enough to dismiss the existing approval.

  

_{Sent by Cursor Automation: Assign PR reviewers}

[ehfeng]

this all seems good, though I wish there was some kind of cmd to run all the examples.

[IlyaasK]

this all seems good, though I wish there was some kind of cmd to run all the examples.

next pr will be ci for this verification purpose

[cursor]

Risk assessment: Low risk.

I assessed the current diff directly. The PR remains limited to documentation and contributor guidance files (.md/.mdx), adding or adjusting Go SDK examples and related example-authoring guidance. There are no runtime code, site config, dependency, workflow, infrastructure, schema, auth/permission enforcement, or model/system prompt changes in the diff.

The latest synchronization after the existing approval only changes .docs/code-example-guide.md and README.md guidance, so it does not increase the risk enough to dismiss the approval. I found no CODEOWNERS file in the repo. Since the PR already has an approved review decision, I’m not adding another approval.

  

_{Sent by Cursor Automation: Assign PR reviewers}

[ehfeng]

interesting, looking forward to seeing how you approach it!

[ehfeng]

i've never seen people ci-build-test code snippets, would be nifty

[IlyaasK]

i've never seen people ci-build-test code snippets, would be nifty

do we want a basic syntac checker or a full on test suite to ensure every code snippet is valid?