{"capabilities":["evidence-bound trust scoring","evidence-admission gating","Sybil damping","honest refusals (insufficient-data / evidence-disputed / unknown-context / invalid-input)","machine-readable reports (report.json) with input_sha256","CLI pipeline integration (check command, exit codes)"],"description":"Deterministic, evidence-bound trust scoring for AI agents — the CWI Verdict Engine v1.0.0 as runnable software. Returns a scored verdict with a full evidence trace, or an honest refusal. Never invents a score. Stdlib-only Python, $0, no network, no accounts.","honest_limits":["scores evidence; does not create trust","insufficient-data is the correct answer for new subjects, not a failure","no telemetry, no network, no accounts — evidence never leaves the machine"],"name":"Verdict CLI","offers":{"currency":"USD","price":0,"status":"store card #5 PENDING — licensing on request, no prices published without approval"},"publisher":{"name":"Cumulative Web Inc","url":"https://cumulativeweb.com"},"requirements":"Python 3.8+, stdlib only, no dependencies, no network","software_version":"1.0.0","source_code":"https://github.com/CumulativeWebInc/cwi-verdict-cli","url":"https://cumulativewebinc.github.io/cwi-verdict-cli/"}