A multi-lens engineering synthesis. Every claim independently checked against its primary source — version-aware — before publication.
Version currency: verified against {{VERSIONS_CHECKED}} as of {{DATE}}.
{{SUMMARY: 3-5 short sentences a busy tech-lead reads in under a minute. Plain words. Lead with the recommendation + the verified fact it rests on, then the main trade-off, then the key risk. Bold the single most important sentence.}}
| Criterion | {{OPTION_A}} | {{OPTION_B}} |
|---|---|---|
| {{CRITERION}} | {{A verdict}} {{T1-3}} | {{B verdict}} {{T1-3}} |
{{FINDING_BODY: plain and direct — what it says, the evidence, and the verified caveat that caps the score. Short sentences; define any term inline. Bold any correction made during verification.}}
{{Why it is contested: the critique, the counter-evidence, the source. Tell the reader not to state it as fact.}}
{{Name the two findings that look like they clash.}}
{{Explain what actually connects them — the link only visible across all the lenses.}}
{{The one-sentence takeaway. If it leans on a weak finding, say so in parentheses.}}
{{What the lenses took as fixed. Name the one perspective that would change the conclusions, not just add nuance.}}
{{The one gap that could flip the findings. Keep it in mind when you build.}}
{{Who this is for + framing: specific moves, not principles.}}
{{Why + exactly how, tied to a finding.}}
{{Why it matters: what it threatens, why no lens addressed it, what answering it unlocks.}}
{{The specific players / versions / patterns in play as of the research date. Name them. State what is current vs deprecated and the version where it changed. Not "the ecosystem is mature" — the actual state.}}
{{The practical core — win or lose "practical" here. Real setup steps, commands, config, minimal code, file layout, env vars. Show the actual path, not "an overview of implementation." Use a code-compare / annotated-diagram / deployment-pipeline widget where it helps.}}
{{Exact languages / frameworks / versions / tools, and WHY each fits the reader's case. Pin versions. A comparison-matrix widget fits here.}}
{{How it connects to what the reader ALREADY runs: APIs, protocols, data formats, compatibility windows. Cite the compat matrix / version-support window. c4-container-view or sequence-flow widget if a picture helps.}}
{{Real numbers with disclosed methodology; the scaling model and where it saturates; the reader's expected load vs that ceiling. Not "scales well." Use metric-bars / range-bar / capacity-scaling-curve.}}
{{Concrete: specific CVEs/GHSA + affected/patched versions, the SPDX license + obligations, the attack surface, the hardening steps. Use the severity-summary widget.}}
{{The decision and why, referencing the reader's stack / team size / scale / constraints BY NAME. State what you'd NOT do and why. weighted-decision-matrix or verdict-card widget.}}
{{Phased, estimable steps (effort S/M/L, sequence, what unblocks what) — concrete enough to hand to the team on Monday. phasing-roadmap or migration-checklist widget.}}
{{Named risks with likelihood/impact and a SPECIFIC mitigation each; the rollback / exit path. No "there are risks" — enumerate them. risk-heatmap widget.}}
{{How claims were checked; the verification tally (N checked / X corrected / Y demoted / Z version-stale); confidence by source tier. The evidence-confidence widget fits here.}}