Strategy and creative research playbook
Establish a brand source of truth
Section titled “Establish a brand source of truth”Use an independent brand card per identity with fields: mission, audience jobs-to-be-done, approved accounts, core proofs and canonical URLs, differentiated promise, tone/voice examples from owned writing, languages, forbidden claims/topics, content rights, CTA destination, and disclosure needs. Do not infer endorsements or achievements.
Suggested brands: personal research/engineering, NovaFabric, KubeIntellect, YazSes. Never quietly post project claims under a personal account unless the owner authorized it.
Attention hypothesis framework
Section titled “Attention hypothesis framework”Every hypothesis must specify: intended platform, audience segment, objective, original evidence asset, hook style, expected mechanism, baseline, time window, measurement source, and stop/adjust criterion. Avoid fictitious algorithm knowledge. Qualitative confidence tags: VERIFIED (first-party/primary evidence), SUPPORTED (credible third-party evidence), HYPOTHESIS (to test).
Select opportunities by scoring 0-3 each: audience pain, novel evidence, clarity/demo potential, timeliness, discussion value, contribution to a real goal. Do not equate raw clickbait reach with success. Document rationale; do not present an arbitrary formula as scientific validation.
Sources of authentic ideas
Section titled “Sources of authentic ideas”- GitHub release notes, merged PRs, reproducible demos, benchmarks with datasets and limitations.
- Accepted/published paper abstracts (state actual publication status), Q&A, learning experiences.
- Genuine questions asked by engineers and users; redact names and private facts.
- Public platform news, standards and technical incidents with primary-source verification.
- Contrasts, before/after stories and tutorials only where both sides are real and comparable.
- Trend opportunities only if they intersect with an existing audience need; no trend spam.
Content engine
Section titled “Content engine”Turn one piece of evidence into: 1 short proof-led LinkedIn narrative, 1 compact technical thread outline, 1 Bluesky/Mastodon note, 1 short-form video script/shot list, 1 visual concept, and 1 FAQ/community answer where welcomed. Do not auto-publish the same copy on every platform. Localize tone and CTA per community, and prefer single topic per post.
Quality self-review: Would the reader gain something useful even without clicking? Does the first sentence state a real claim? Is proof visible? Is the CTA optional rather than manipulative? Would the author be comfortable publicly defending the post? Is the claim true in the latest source revision?
Calendar discipline
Section titled “Calendar discipline”Start with 2 weeks, 2-4 evidence-backed posts/week on the best-fit channels if resources permit; this is a planning heuristic, not an algorithmic requirement. Batch creation, but retain a buffer for fact-checking and replies. No automatic weekend/weekday formula without audience evidence.
Community and reputation
Section titled “Community and reputation”Engage as a real person. Offer useful answers and thoughtful replies, not copy-paste promotion. Explicitly respect each subreddit/community’s self-promotion rules. Never automate unsolicited private messages or retaliatory response. Escalate legal/medical/security/crisis topics to the human owner.
