AINL_INFRASTRUCTURE_DIAGNOSTIC.md

AINL Infrastructure Diagnostic Report — DEPRECATED

2026-05-19 audit: This file was deprecated as part of the credibility/positioning pass tracked in docs/competitive/LONG_TERM_FIXES_TRACKER.md row T1.11.

Why: The original report mixed real operator signals (17 active cron jobs on the project author's stack, 99.7% uptime metric) with hypothetical "traditional agent loop" baseline comparisons ($6.03/day orchestration vs $0.97/day AINL → $180.90/month savings). The same "7.2×" figure appeared here, in AINL_COST_SAVINGS_REPORT.md, and in AINL_OPERATIONAL_DEPLOYMENT_REPORT.md — each with slightly different framing — which made it easy to over-quote.

The original document is preserved unchanged at archive/legacy_savings_reports/AINL_INFRASTRUCTURE_DIAGNOSTIC.md for provenance.

Where to look instead

What you wanted	Where it lives now
Reproducible orchestration-token benchmarks	docs/CLAIMS_AND_EVIDENCE.md §1 + scripts/benchmark_compile_once_run_many.py
Operator field signals (Class b evidence)	docs/competitive/PRODUCTION_EVIDENCE.md Cases 1–2
Real uptime / reliability metrics	STATUS.yaml (auto-refreshed by scripts/refresh_repo_stats.py); execution reliability percentages in tooling/benchmark_runtime_results.json
Honest baseline filter	docs/competitive/WHEN_AINL_DOES_NOT_HELP.md

Going forward: infrastructure diagnostics that include cost claims must source every numeric value from a committed JSON under tooling/ (with the baseline tag attached) or be explicitly labeled as operator-specific scenario math.