Description
vettd's only report outputs today are machine-readable JSON and colorized terminal text. There's no standalone, shareable report artifact a human can open or attach to a build without a terminal. A competitive scan against another AI-artifact scanner (stoa) showed this gap concretely: stoa's --html output produces a self-contained styled report (summary cards, per-finding detail, severity badges) suitable for archiving as a CI artifact or sharing with a reviewer; vettd has no equivalent.
Scope
- A new output flag (e.g.
--html <FILE>) on the folder/repo/file scan subcommands that renders a self-contained static HTML report from the same finding data already produced for JSON output.
- Report should surface the same severity/risk/reasoning detail currently only visible in terminal output (risk score breakdown, capability tags, contributing signals) in a browsable form.
Acceptance Criteria
Description
vettd's only report outputs today are machine-readable JSON and colorized terminal text. There's no standalone, shareable report artifact a human can open or attach to a build without a terminal. A competitive scan against another AI-artifact scanner (stoa) showed this gap concretely: stoa's
--htmloutput produces a self-contained styled report (summary cards, per-finding detail, severity badges) suitable for archiving as a CI artifact or sharing with a reviewer; vettd has no equivalent.Scope
--html <FILE>) on the folder/repo/file scan subcommands that renders a self-contained static HTML report from the same finding data already produced for JSON output.Acceptance Criteria
--html <FILE>writes a self-contained (no external asset fetches) HTML reportscan folder,scan repo, andscan file