distribution-metrics

name	distribution-metrics
description	Track and summarize local publishing-channel metrics for Algo Mind essays. Use when the user wants to learn from distribution results, create UTM links, record published channel posts, capture manual metric snapshots, compare hooks/assets/channels, or review what worked after using the go-to-market skill.

Distribution Metrics for Algo Mind

Track channel results locally after an essay is published and distributed. This skill owns the Learn stage: create tracked links, record where the essay was posted, add manual snapshots, and summarize what worked.

Keep this local-first. Do not introduce dashboards, API keys, browser scraping, or committed private analytics unless the user explicitly asks.

Local Data

Metrics are written to dist/distribution-metrics/metrics.jsonl, which is ignored by git through the repo's existing dist/ rule.

Each record is append-only JSONL. There are two record types:

• publish — one platform post, with URL/id, tracked URL, asset type, and hook type.
• snapshot — one metric snapshot for that platform post.

Workflow

1. Create a UTM-tracked blog URL for the channel/post variant.
2. Record the platform post after publishing.
3. Add snapshots around 1h, 24h, 72h, 7d, and 30d.
4. Summarize by essay, channel, hook, and asset.

Commands

Create a tracked blog URL:

node .claude/skills/distribution-metrics/scripts/make-tracked-url.mjs <essay-slug> <channel> --content <variant>

Record a published post:

node .claude/skills/distribution-metrics/scripts/record-publish.mjs <essay-slug> <channel> --url <platform-url> --tracked-url <utm-url> --asset <asset-type> --hook <hook-type>

Record a manual metrics snapshot:

node .claude/skills/distribution-metrics/scripts/record-snapshot.mjs <essay-slug> <channel> --url <platform-url> --impressions <n> --likes <n> --comments <n>

Summarize:

node .claude/skills/distribution-metrics/scripts/summarize-metrics.mjs --slug <essay-slug>

References

• references/local-metrics-playbook.md — metric fields, snapshot cadence, channel-specific manual fields, and learning questions.

Notes

• This skill is canonical at .claude/skills/distribution-metrics/ and symlinked to .codex/skills/distribution-metrics for Codex.
• Pair this with go-to-market: use go-to-market to package and publish, then use distribution-metrics to learn from the result.