Global distribution operating standard

Review date: 2026-10-08 | Status: Research and offline proposals; not a production posting or first-party analytics optimizer.

India-focused research extension (2026-10-15)

Nine candidate India-origin networks, media/creator surfaces and community systems are now classified in the regional catalog. They are manual-only/unverified; Kutumb, LocalCircles and Arattai Channels are human-only and require group/owner permission or opted-in followers. A separate Zoho Flow integration for Arattai does not authorize a ReachWeaver publisher. Koo is not active and is intentionally excluded. Preserve per-platform evidence dates and the India TikTok restriction review.

Dated regional research (2026-10-14)

The regional channel expansion adds 27 manual-only surfaces and South Korea, Taiwan, Thailand and Russia pilots. reviewed_on is the catalog edit date; each channel’s checked_on retains its own actual evidence-check date. A research_source homepage is not proof of native scheduling, API rights or account authorization. Messaging/chat communities remain human-only; the India pilot omits TikTok until current availability and law are checked.

Authority and source of truth

1. reachweaver/distribution_registry.json is the versioned, testable snapshot of channel identities, scheduling claims, integration reality, market and goal-specific editorial priorities, and experimental time windows. This file is not proof of proprietary algorithm knowledge or a future traffic forecast.
2. reachweaver/distribution.py validates and reads that file and can generate offline channel research using reachweaver distribution-plan. Every channel in a plan carries its checked_on date and a fresh/aging (more than 60 days)/stale (more than 120 days) label; reachweaver registry-age --registry distribution lists the facts to re-verify (policy). The same files are shipped in the standalone Claude Code skill. Changes must pass tests/test_distribution.py and mirror checks.
3. docs/GLOBAL-DISTRIBUTION-2026-10-08.md explains audience strategy, local-market ethics, moderation and pilots. docs/PLATFORM-ALGORITHMS-SCHEDULING-2026-10-08.md explains official algorithm guidance and native scheduler distinctions. docs/OPEN-SOURCE-DEVELOPER-GROWTH-PLAYBOOK.md treats long-term contributor conversion.
4. docs/PLATFORM-MATRIX.md is the human-readable view of the live ReachWeaver delivery adapter status. It mirrors reachweaver_adapter in reachweaver/platform_registry.json and the scheduling modes here, and tests/test_platform_matrix_docs.py fails when they disagree. If scheduling and API status disagree, do not infer a new publishing capability; the most conservative statement prevails. No real accounts have been connected or live end-to-end publishing verified.
5. Skill references are portable distilled operating rules, not an independent source of supposedly verified platform capabilities. Update them when platform or moderation rules change.

Use it without publishing

# All commands below are local/read-only, require no social login, and do not schedule any post.
python3 -m reachweaver distribution-plan --region india --goal adoption --limit 5
python3 -m reachweaver distribution-plan --region china --goal contributors --limit 6 --json
python3 -m reachweaver distribution-plan --region nigeria --goal collaboration --date 2026-10-14 --json
# Worldwide has no single audience timezone: supply a specific one for any time experiment.
python3 -m reachweaver distribution-plan --region global --goal awareness --date 2026-10-14 --timezone Europe/Rome --json

From a copied Claude Code skill, use python3 ~/.claude/skills/reachweaver/scripts/run.py distribution-plan ... while working in the target project.

The result explicitly says OFFLINE_RESEARCH_PROPOSAL_NOT_SCHEDULED. The native_schedule field describes a platform or account role; reachweaver_delivery says what our code has. The illustrative_time_tests field is an untested local-time hypothesis, with its IANA zone converted to an offset-aware UTC instant. Neither field schedules a remote platform post, confirms an account, recommends a guaranteed good hour or grants publishing consent.

Change-management contract for platform and market research

For every new/changed factual claim:

1. Prefer current primary sources: platform policies/help/docs, product controls and API specs. Keep a https:// URL, observation date, specific relevant account role, eligibility, format limits, and difference between native UI and API. If a source becomes stale/unavailable, mark unverified rather than retaining a strong claim.
2. Separate (a) platform-native scheduling, (b) permitted external API/client scheduling, (c) currently mock-tested ReachWeaver adapters, (d) live verified owner account and real remote receipts. A provider announcement never proves (d).
3. Record market-priority changes as editorial hypotheses; explain target persona, actual problem, language and first-party experiment that could falsify the selection. No invented ranking weights, growth guarantees or universal optimal times.
4. Country and language are separate: localized user documentation or captions do not establish speech recognition, hardware or OS language support. Obtain human local-language QA, use correct character sets, and label unsupported features transparently. Check local account-creation requirements and legal/regional availability; do not bypass platform verification.
5. Every modified market must reference defined platform IDs, all four supported goals, an IANA timezone for country-level time proposals, and no duplicated priority IDs. Update both portable and canonical registries with python3 scripts/sync_skill.py.
6. Run python3 -m unittest discover -s tests -v, inspect the PR diff, and let GitHub CI verify Python 3.11, 3.12, 3.13 and 3.14. Changes to writing/publishing permissions require security review and a separately approved live pilot.
7. Recheck policies before each campaign; dates in a versioned file indicate the last document review, not a guarantee that every cited product feature still exists.

Community-specific rules

• Hacker News: An exceptional human_only channel. Research and troubleshooting checklists are allowed, but ReachWeaver must never write, polish, translate, automate or schedule HN-facing titles, submissions or comments. See HN official guidelines, Show HN, current restriction, and 2026 moderator note. Do not suggest republishing to circumvent flags or coordinated votes.
• Reddit: Native scheduled posts are a moderator capability, not a regular-user broadcasting feature (Reddit help). Regular promotion is a rule-by-rule manual community decision; avoid repeated mass linking.
• Facebook Groups, Discord and independent communities: Permission and local rules outrank growth opportunity. Membership does not confer promotional broadcast authority; no mass unsolicited messaging.
• Chinese/regional platforms: Do not claim permission, API access or local availability before account-role and regulation checks. Use human QA for Simplified Chinese, Japanese, Arabic, Swahili, French, Portuguese, Hindi and other languages as appropriate.

Evidence ladder for existing activity

Use examples/yazses-channel-audit.example.json to distinguish owner_reported from verified activity. Upgrade only with the real URL, timestamp, source account, first-party receipt and, if measured, a native export with clear metric definition. Never commit credentials, private analytics, login cookies, contact lists or third-party personal information.

The channel audit is illustrative and NOT wired into the send queue or account config. Publicly posting content elsewhere does not consent to ReachWeaver posting.

Sample four-week regional experiment

Week	Activity	Gate	Observable useful result
1	Verify current LinkedIn, DEV, YouTube, GitHub and Snapcraft presence; fix demo-to-install funnel	Owner URLs and public feature validation	Native click sources, useful repo/docs journey
2	One original offline Linux workflow demo, distinctly adapted to YouTube, DEV and LinkedIn	Human proof/content review; site rules	Watched demos and relevant troubleshooting questions
3	India tutorial and one China developer/video trial if authenticated and human-localized	Correct account eligibility; language QA; no false Chinese dictation promise	Country-specific qualified site/repo referrals
4	One Nigeria/Kenya/South Africa and one Japan/Indonesia test	Relevant community/locale, no broad unsolicited broadcasts	Qualified visits, installations, issue authors, time cost and moderation outcomes

For each post and locale, log content_evidence, published_url, platform_account, market/language, manual/native/API, local+UTC time, creative_format, approval, moderation_outcome, views_definition, +24h/+7d metrics, qualified referrals, installs, contributors, and missing_data. These are future measurement requirements, not fields automatically collected by the current planner.

Metrics rule: Platform views, impressions, unique people, clicks and repository stars are different denominators. Do not add them into a single fabricated global audience. Prefer qualified acquisition and repeat contribution, with honest uncertainty, to large but irrelevant reach.