| 1 | # Skill: Press-Developer Convergence Detection |
| 2 | |
| 3 | ## Purpose |
| 4 | Identify weeks where press/industry coverage and developer activity converge on the |
| 5 | same theme, surfacing non-obvious editorial angles from the correlation. |
| 6 | |
| 7 | ## Method |
| 8 | |
| 9 | ### 1. Correlation Scan |
| 10 | Compare the week's press themes (from compact press-context artifact) against: |
| 11 | - New repo categories showing unusual growth |
| 12 | - Trending repos with thematic overlap to press narratives |
| 13 | - Community reaction repos (alternatives, self-hosted replacements, protest forks) |
| 14 | |
| 15 | ### 2. Convergence Strength Tiers |
| 16 | |
| 17 | | Tier | Definition | Example | |
| 18 | |------|-----------|---------| |
| 19 | | **Strong** | Same-week developer repo response directly caused by press event | Copilot billing article → self-hosted AI workspace repos spike (W23) | |
| 20 | | **Meta** | Press narrative mirrors a pattern visible in developer metrics | Inflated AI ARR coverage ∥ star-inflation campaigns (W22) | |
| 21 | | **Weak** | Category overlap without causal link | General AI coverage + general AI repos trending | |
| 22 | |
| 23 | ### 3. Editorial Application |
| 24 | - **Strong convergence**: Lead with the narrative; developer data validates/extends press |
| 25 | - **Meta convergence**: Use as a thematic bridge; highlight the irony or structural parallel |
| 26 | - **Weak convergence**: Mention only if no stronger angle exists; caveat the correlation |
| 27 | |
| 28 | ## Reusable Patterns |
| 29 | - Platform billing/policy changes → monitor for self-hosting spikes within 7 days |
| 30 | - Acquisition announcements → watch for fork/migration activity in affected ecosystems |
| 31 | - Security disclosures → track patching velocity and alternative-tool adoption |
| 32 | - Regulatory announcements → observe compliance-tooling repo creation |
| 33 | |
| 34 | ## Quality Gate |
| 35 | - Never present correlation as causation without temporal evidence |
| 36 | - Always cite both the press source (URL, date) and the repo evidence (crawl data) |
| 37 | - Distinguish press-first (coverage preceded developer activity) from developer-first |