promote

Generate data-driven X/Twitter posts for plugin releases with ASCII cards ready for CodeSnap screenshots.

When to use

Run /promote vX.Y.Z after tagging a release. Optionally add a title: /promote v2.8.0 "LiveView Streams Overhaul".

Output

All files go to scratchpad/x-posts/{version}-release.md (create directory if needed).

The output file contains three sections:

1. ASCII Card — box-drawing table for CodeSnap screenshot
2. Tweet Thread — 3-5 tweets in proven format
3. CodeSnap Command — exact CLI command to render the card
4. Fact-Check — every claim with verifiable data source

Execution Flow

Step 1: Gather Data

Collect these metrics (all are verifiable commands — run them, don't estimate):

Metric	How to get it
File count + insertions/deletions	git diff --stat {prev-tag}..{tag}
Skill count	ls plugins/elixir-phoenix/skills/ | wc -l
Agent count	ls plugins/elixir-phoenix/agents/*.md | wc -l
Iron Law count	Grep for numbered Iron Laws in CLAUDE.md
Eval scores	make eval-all (run it, report actual numbers)
Changelog entries	Read CHANGELOG.md for this version's section
Before/after metrics	Depends on release — look for quantifiable changes

The strongest posts have before/after comparisons. Look for changes in CHANGELOG.md that have measurable deltas (line counts, counts of affected files, error rates, coverage numbers).

Step 2: Write the ASCII Card

The card is the visual centerpiece — it gets screenshotted via CodeSnap and attached to tweet 1.

Rules:

• Exactly 72 characters wide (including borders) — consistent with past posts
• Use Unicode box-drawing: ┌─┐│└┘╞═╡╤╪╧ for borders and separators
• Include: version + title, headline stats, before/after table, repo URL
• The table content comes from Step 1 data — pick the 3-5 most impressive changes
• VERIFY ALIGNMENT: After writing the table, run a Python script to check every line has the same visual width. Off-by-one errors are the most common problem — em dashes (—) and arrows (→) are single-width but easy to miscount. Every line between the top ┌ and bottom └ borders must be exactly 72 visual characters

Also save the ASCII card as a separate .txt file at scratchpad/x-posts/{version}-table.txt for CodeSnap input.

See ${CLAUDE_SKILL_DIR}/references/templates.md for the proven card format and past examples.

Step 3: Write the Tweet Thread

Follow the hook → findings → details → CTA structure. Read ${CLAUDE_SKILL_DIR}/references/templates.md for proven tweet patterns.

Thread rules (from analytics on 7 posts, 6.9K-9.9K views):

• Tweet 1 (Hook): Version + one compelling sentence about what changed. Include headline stats (files changed, skills, agents). Repo link here — not in a reply. Add #ElixirLang #ClaudeCode hashtags.
• Tweet 2-3 (Findings): The specific changes with numbers. Use numbered lists. Before/after metrics are the highest-performing content type.
• Tweet 4-5 (Details/CTA): Additional changes + soft CTA ("Try it and let me know how it feels"). Link to release page.
• Max 5 tweets — casual readers drop off after 4-6. If you have more content, cut the least impactful items.

Voice and tone:

• Data-heavy, transparent, no-hype — this is the proven #1 strength
• Write like a builder sharing real numbers, not a marketer
• Concrete numbers over adjectives ("32/40 descriptions rewritten" not "improved many descriptions")
• Never mention API costs or token prices — subscription user, cost per call is irrelevant

Step 4: CodeSnap Command

Output the exact command to render the ASCII card as an image. The card .txt file from Step 2 is the input.

Use the config file at scratchpad/x-posts/codesnap-claude-dark.json (dark gradient, tight margins, no watermark):

codesnap -f scratchpad/x-posts/{version}-table.txt \
  -o scratchpad/x-posts/{version}-card.png \
  --config scratchpad/x-posts/codesnap-claude-dark.json \
  --title "claude-elixir-phoenix {version}" \
  -l text

Then render and verify the image looks correct by reading the output PNG.

Step 5: Fact-Check

Every claim in the tweets must have a corresponding entry in the fact-check section:

## Fact-Check

- File stats: `git diff --stat {prev}..{tag}` (output: N files, +X/-Y)
- Eval scores: `make eval-all` output — N skills avg X.XXX, N agents X.XXX
- [claim]: [exact command or file:line that proves it]

If a claim can't be verified with a command or file reference, flag it as UNVERIFIED and suggest rewording.

Bundling Strategy

When deciding whether a release warrants its own post:

• Standalone post: Major features, impressive before/after metrics, new capabilities
• Bundle with next release: Bug fixes, description tweaks, minor config changes
• Skip entirely: Patch releases with <5 files changed and no user-facing changes

The v2.6.1 post (37 likes, 1.4K views) vs v2.6.0 (150 likes, 5.5K views) demonstrates this — back-to-back releases dilute signal. When in doubt, wait and bundle.