Concept-popularization epic: make the ideas known and attractive

Filed 2026-07-02. The existing marketing surface (AEO-MARKETING-NEXT-STEPS, internal/marketing/aeo.go) is mechanism — answer-engine feeds, schema.org, SEO front-matter. This epic is the human-facing half: the artifacts, stories, demos, and moves that make a person *want* the concepts and *remember* them. The two are complements — AEO makes fak findable by machines; this makes it compelling to people.

The five concepts we are popularizing

Every ticket below serves at least one. When authoring, name which.

  1. Treat the tool call like a syscall — the model proposes, the kernel disposes. The one-line mental model that makes the whole design click.
  2. Verify, don’t trust (DOS) — a false “done” is refused from git evidence, not the agent’s word; refusals carry a structured reason from a closed vocabulary; recalled memory is re-verified at read time. The trust substrate under a fleet of agents.
  3. The addressable, bit-exact KV cache — reach into the middle of a kept run, evict one span, leave the cache bit-for-bit identical (max|Δ| = 0); long sessions stop getting expensive because the cached prefix survives.
  4. Default-deny capability gate + prompt-injection quarantine — refusing an irreversible action does not depend on catching an attack; the lever was never wired. Suspicious tool results are held out of context by structure, not by a classifier.
  5. One static Go binary, drop-in in front of any agent — repoint one base URL; the same artifact a dev runs on a laptop is what a platform team hardens for a fleet.

The eleven dimensions (each ticket lands in exactly one)

Dim Name The question it answers
A Explainer content “I read one page and now I get it.”
B Visual & diagram assets “One picture and I can re-explain it to a colleague.”
C Interactive & runnable demos “I ran it in 60 seconds and saw the verdict.”
D Positioning & comparison “How is this different from the thing I already use?”
E Social proof & community “Other people I respect use and vouch for this.”
F Developer experience & onramp “Nothing stood between me and my first win.”
G Integration recipes “It dropped into the exact stack I already run.”
H Benchmark-as-story “The number is real, sourced, and I feel its stakes.”
I Memorable framing & naming “I can repeat the pitch from memory a week later.”
J Distribution & channels “It showed up where I already spend attention.”
K Adoption measurement “We can tell whether any of this is actually working.”

Rules for every ticket (so each is a self-contained unit of work)

The 50 tickets

The full enumerated list with per-ticket bodies is filed as GitHub issues under the popularization label. This doc is the durable framing; the issues are the units of work. Dimension coverage target: A×8, B×5, C×6, D×5, E×5, F×5, G×4, H×4, I×3, J×3, K×2.

Verify

gh issue list --label popularization --state open        # the 50 units of work
python tools/seo_aeo_scorecard.py                        # new docs don't red SEO
python tools/popularization_readiness_scorecard.py       # dim K: does the front door convert?

The last command is the epic’s own self-verifying number: the dimension-K popularization-readiness scorecard models a first-time visitor’s funnel (land → orient → trust → install → act) and drives popularization_debt toward 0 — a missing hero command or an unreachable adoption doc raises debt, so the epic ratchets itself down over time and folds into tools/scorecard_control_pane.py.