Skip to content

feat(apex): lead the hero with why, not the paste command#559

Merged
isuttell merged 1 commit into
mainfrom
apex-why-first-copy
Jun 17, 2026
Merged

feat(apex): lead the hero with why, not the paste command#559
isuttell merged 1 commit into
mainfrom
apex-why-first-copy

Conversation

@isuttell

Copy link
Copy Markdown
Contributor

Summary

A real visitor bounced off the live apex landing site saying "I must not be in the industry. I just don't understand what this is for." The hero led with a how-to instruction ("Paste this into a shell-capable agent. It installs agent-paste...") plus a second imperative restating the action, so two HOW sentences stacked before the visitor was ever told WHAT the product is or WHY they'd want it. The headline ("Open it anywhere") promised value; the subhead immediately talked about shells and JavaScript. This rewrite leads with the human pain and payoff, then routes the reader into the copy-prompt funnel. The marketing page is human-focused; agents read /agents.md and /llms.txt, not this page.

Diagnosed and validated with a 3-lens adversarial review (cold visitor, ICP developer, brand enforcer) grounded in YC's website-clarity guidance. All three originally ranked the live hero last; all three pass the rewrite.

Changes

  • Hero lead rewritten human-first: "Your AI agent built a report, a dashboard, a prototype, and it is trapped in a chat window. agent-paste turns it into a link you can open in any browser and send to anyone. Tell your agent to publish it; you get the link."
  • Dropped the redundant HERO.heroAction "Copy the prompt... / See the session" directive.
  • Added a "When you'd reach for it" use-cases block under the demo (concrete jobs, each ending in the link).
  • Reordered the detail pane so mechanism (commands, MCP) and the reasons-to-believe sit below the fold; extended the reveal stagger to .d7/.d8.
  • Reframed the four feature titles from mechanism to benefit; dual-audience detail stays in the body as a human payoff.
  • Aligned META_DESCRIPTION (search snippet) to the human-first framing, under the 125-char crawler cap.
  • De-flated the static example page: stopped captioning it "static demo artifact, not a live page".
  • Filed deferred follow-ups in docs/ops/apex-landing-todo.md (live feat-of-strength, fabricated demo link, headline-word A/B, banned-token test hardening), linked from project-status.

Risk: LOW

  • Areas touched: apex marketing copy + page structure + one static example page + ops docs. No app logic.
  • Security: none.
  • Performance: none (static prerender).
  • Breaking: none.

Test plan

  • pnpm verify green
  • pnpm test:coverage green (CI Validate gate)
  • apex render tests pass (meta-length cap, clipboard targets, em-dash ban hold)
  • ziw-code-review: clean, no findings, READY FOR PR
  • Deployed to preview (preview.agent-paste.sh): new hero served, old copy gone, use-cases block present, block order correct, example page de-flated

Issue: AP-374

@coderabbitai ignore

A visitor bounced saying "I just don't understand what this is for."
The hero led with a how-to instruction ("Paste this into a shell-capable
agent...") before answering what/why, and a second imperative restated
the action. Rewrite the lead to a human-first what/why ("trapped in a
chat window" -> "a link you can open and send to anyone"), drop the
redundant heroAction directive, add a "When you'd reach for it" use-cases
block under the demo, reorder the detail pane so mechanism and the
reasons-to-believe sit below the fold, reframe the four feature titles to
benefits, align the search-snippet meta description, and stop the static
example page from captioning itself as fake.

Validated with a 3-lens adversarial review (cold visitor, ICP dev, brand
enforcer); all three pass the rewrite. Deferred follow-ups in
docs/ops/apex-landing-todo.md.

Issue: AP-374

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

@cursor cursor Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Risk: simple
Decision: approve

Ticket triage

  • Intended change: Rewrite the apex landing hero to answer what/why before mechanism (human pain → shareable link), add a "When you'd reach for it" use-cases block, reorder detail-pane content, reframe feature titles to benefits, align META_DESCRIPTION, and de-flate the static example page caption.
  • Scope match: Yes. All in-scope items from AP-374 are present; deferred follow-ups are documented in docs/ops/apex-landing-todo.md.

Review findings

  • Blocking: None.
  • Non-blocking:
    • Fabricated demo Access Link remains (pre-existing; filed as follow-up #2 in apex-landing-todo.md).
    • Banned-token test hardening deferred (follow-up #4); acceptable for this copy-only change.
    • No structural test for the new use-cases block — consistent with render.test.tsx policy of asserting contracts, not marketing prose.

Merge checklist

  • Ticket linked: ✅ AP-374
  • Scope matches: ✅
  • Checks green: ✅ Validate, CodeQL, Secret scan, Postgres smoke
  • Tests/docs appropriate: ✅ Render contract tests + ops follow-up ledger
  • No blocking findings: ✅
  • No high-risk areas: ✅ Marketing copy/static HTML only
  • Merge-safe: ✅

Summary

Clean, focused marketing rewrite. heroAction removal is safe — the demo block still carries the copy-prompt funnel below the fold. META_DESCRIPTION is 110 chars (under the 125-char cap). Reveal stagger extended correctly for the new block order. Approve for merge.

Open in Web View Automation 

Sent by Cursor Automation: First Pass PR Reviewer

@isuttell isuttell merged commit 13e8475 into main Jun 17, 2026
11 checks passed
@isuttell isuttell deleted the apex-why-first-copy branch June 17, 2026 23:01
@github-actions

Copy link
Copy Markdown

agent-paste PR preview resources were cleaned up. The shared Preview GitHub Environment is retained for future preview deploys.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant