Vision Lock v1: AI News Aggregator -> AI Enthusiast Daily (ADR-0007)

- docs/vision/VISION_LOCK_V1.md: permanent editorial mission, audience (builders/tinkerers/
  local-AI users/home-lab/agent devs), Editorial Test ('what would an enthusiast DO?'),
  architectual constraint (editorial vision is primary system), priority order.
- docs/vision/ADR-0007_EDITORIAL_PIVOT.md: accepted pivot record, before/after table,
  Sprint 1 evidence (high-signal = actionable, low-signal = spectator content).
- docs/vision/EDITORIAL_GUIDELINES.md: operating rules; score is draft, taste overrides;
  hand-before-machine; memory must store 'what can be done' not 'what happened'.
- docs/vision/SECTION_DEFINITIONS.md: 5 sections (What Shipped / Run It Locally /
  Benchmarks & Builds / Problem Solved / Worth Trying Tonight).
- docs/vision/SPRINT_1_FINDINGS.md: reference evidence, 45/200 UNCATEGORIZED taxonomy gap.
- Issue_001.md: handcrafted prototype edition from Sprint 1 stories; benchmark for all future automation.

This is the architectural directive. No scoring/narrative/memory work proceeds except in
service of the Vision Lock. Sprint 2 (Lens) stays blocked pending manual review tallies.

docs/vision/SECTION_DEFINITIONS.md

@@ -0,0 +1,35 @@
+# SECTION DEFINITIONS — AI Enthusiast Daily
+
+The publication is organized into five working sections. Every published story maps to exactly one. These are editorial sections, not the Sprint 1 classifier buckets (those are a separate, internal taxonomy — see athena_review_report.txt).
+
+---
+
+## 1. What Shipped Today
+**Things people built.**
+New open-source tools, models released with weights, agents shipped, MCP servers, GitHub projects that a reader can go use right now.
+*Editorial test:* "Can I install or run it today?" → Yes = in.
+
+## 2. Run It Locally
+**Local models, GGUF, Ollama, inference, benchmarks on consumer hardware.**
+Anything about running AI on your own machine — quantization, VRAM budgets, llama.cpp tricks, Mac/Linux/RTX setups, home-lab builds.
+*Editorial test:* "Could I replicate this on hardware I own?" → Yes = in.
+
+## 3. Benchmarks & Builds
+**Hardware and performance experimentation.**
+Head-to-head GPU comparisons, concurrency tests, latency/throughput numbers, quantization-quality showdowns, build logs with measurable outcomes.
+*Editorial test:* "Are there real numbers I can compare against my own setup?" → Yes = in.
+
+## 4. Problem Solved
+**Real-world outcomes and workflows.**
+Production war stories, failure post-mortems, "we fixed X and here's the method," agent-identity/permissioning, cost-reduction playbooks.
+*Editorial test:* "Did someone actually ship something and learn a transferable lesson?" → Yes = in.
+
+## 5. Worth Trying Tonight
+**High-action recommendations.**
+The short list — 3 to 5 items max — that a reader could act on *this evening*. Curated, not comprehensive. The "if you only do one thing" section.
+*Editorial test:* "Would a builder drop what they're doing and try this tonight?" → Yes = in.
+
+---
+
+## Mapping note (Sprint 1 → sections)
+The internal classifier buckets (LOCAL AI, PROBLEM SOLVED, RESEARCH, INFRASTRUCTURE, SHIPPING, MODEL RELEASE, BUSINESS, CULTURE, UNCATEGORIZED) feed these sections but are NOT identical to them. Section assignment is an editorial act, not a mechanical copy. A RESEARCH paper with a clear builder angle may land in Benchmarks & Builds; a MODEL RELEASE with no reproducible substance may be rejected entirely.