synthesis-content-framing · v2.0.0 · 2026-10-06 · sha256 a957b2d4f8324bd5
synthesis-content-framing v2.0.0A
Immutable. This exact content never changes and is served at /api/v1/blob/a957b2d4f8324bd5.
--- name: synthesis-content-framing description: "Frame synthesis engineering and synthesis coding articles: topic, sophistication, engagement and confidentiality gates; attribution and terminology rules. Use when writing synthesis engineering content, framing technical articles, checking content quality gates, or publishing thought leadership." license: "CC0-1.0" depends_on: [] metadata: author: "Synthesis Engineering" version: "2.0.0" format: v5 source_repo: "github.com/synthesisengineering/synthesis-skills" source_type: "public" --- # Synthesis Content Framing A methodology for writing and publishing synthesis engineering and synthesis coding thought leadership. This skill covers topic selection, sophistication standards, engagement patterns, quality gates, and operational rules. ## Binding rules Rules 1 to 4 are the four blocking gates, numbered as they are everywhere else. If any gate fails, stop and fix it before any other check. 1. **Gate 1, Topic:** the article is about building software. Content editing and publishing workflows fail, because the output of the work must be software or code. 2. **Gate 2, Sophistication:** the example justifies AI collaboration (10K+ lines, 50+ files, 3+ services) and would impress engineers at top AI labs or Fortune 500 CTOs. 3. **Gate 3, Engagement:** open with narrative tension, earn the insights, use vivid specifics, and pass the share, memory and feeling tests. 4. **Gate 4, Confidentiality:** no client names, business strategy or internal positioning language. 5. **The safety rules are automatic rejections:** no article confirms a real security or confidentiality incident or reads as a confession of one. A deleted article is gone only when its live URL returns 404. 6. **The human directs; the AI executes.** Every "I" statement says who acted, and each error goes to whoever made it; the human's usual error is accepting AI output without verifying it. 7. **No fake human collaborators.** Every human mentioned in an article is a real person. 8. **Anonymize clients, advisory relationships and identifying company or project names; name and link open source projects.** Real, verifiable examples build credibility. 9. **Frame AI as assisting human-directed work,** never as replacing human judgment, generating journalism or deciding on its own. Public perception turns on the distinction. 10. **Write "synthesis coding" and "synthesis engineering" in lowercase in prose,** in title case only in headlines, and never hyphenated. 11. **Link by dependency, not by quota.** synthesis-reader-briefing's series-dependency contract decides whether an article needs a prerequisite link; a standalone article needs none. 12. **Pair a concrete example with the pattern it teaches and how to apply it,** and name the primary audience in the first or second paragraph. ## Contents - [references/principles.md](references/principles.md): the public principles: the core principle, the expert operator's toolkit, the direction dynamic, the five pillars, what synthesis engineering is not, and how synthesis coding compares with other approaches. Read it when an article explains or argues for the discipline. - [references/operational-rules.md](references/operational-rules.md): the mechanics of content creation: attribution, scope, confidentiality, perception, audience, cross-linking, terminology, update notes, series integration, the CC0 notice. Read it when drafting or revising an article. - [references/quality-gates.md](references/quality-gates.md): the full gate criteria, safety rules, quality standards, pre-publish checklist and examples of correct framing. Read it when checking an article against the gates, and before publishing. - [references/coverage-map.md](references/coverage-map.md): where each part of the 1.1.0 text now lives. - Quality Gates Summary, Related: below. ## Quality Gates Summary Every article must pass four blocking gates before publication. If any gate fails, stop and fix before proceeding. 1. **Gate 1: Topic Gate** — Is this article about building software? Not content editing, not publishing workflows. The output of the work must be software/code. 2. **Gate 2: Sophistication Gate** — Is the example impressive enough to justify AI collaboration? Minimum scale thresholds apply (10K+ lines, 50+ files, 3+ services). Would this impress engineers at top AI labs or Fortune 500 CTOs? 3. **Gate 3: Engagement Gate** — Is this compelling reading? Does it open with narrative tension, contain earned insights, include vivid specificity, and pass the share/memory/feeling tests? 4. **Gate 4: Confidentiality Gate** — Does this protect all private information? No client names, no business strategy, no internal positioning language. Beyond gates, articles must also pass safety rules (incident confirmation, confession detection), quality standards (content value, example ethics), and operational checks (attribution, voice, cross-linking). For full gate criteria, checklists, and examples, see: `references/quality-gates.md` ## Related Part of the [synthesis writing](https://synthesiswriting.org) craft — the writer writes, the AI assists.