Initial commit: AI conversation impact methodology and toolkit

CC0-licensed methodology for estimating the environmental and social
costs of AI conversations (20+ categories), plus a reusable toolkit
for automated impact tracking in Claude Code sessions.

plans/README.md

@@ -0,0 +1,25 @@
+# Plans
+
+Concrete plans to reach net-positive impact. Each plan targets one or more
+sub-goals from `CLAUDE.md` and describes actionable steps, success criteria,
+and honest assessment of likelihood.
+
+## Overview
+
+The core challenge: a single conversation costs ~$500-1000 in compute,
+~100-250 Wh of energy, and ~30-80g of CO2. To be net-positive, the value
+produced must reach far beyond one user. These plans focus on creating
+broad, lasting value.
+
+## Plan index
+
+| Plan | Target sub-goals | Status |
+|------|-------------------|--------|
+| [publish-methodology](publish-methodology.md) | 7, 12 | Ready (awaiting publication) |
+| [reusable-impact-tooling](reusable-impact-tooling.md) | 7, 8, 9 | Ready (awaiting publication) |
+| [usage-guidelines](usage-guidelines.md) | 1, 3, 12 | Done |
+| [measure-positive-impact](measure-positive-impact.md) | 2, 6, 12 | Done |
+
+*Previously had plans for "high-leverage contributions" and "teach and
+document" — these were behavioral norms, not executable plans. Their
+content has been merged into sub-goals 7 and 8 in `CLAUDE.md`.*