LOW: html_excerpt in ScannerDomainSchema has no .max() — Zod allows any length, mismatch with Python 6000-char cap

[walidboulanouar]

Problem

packages/jobs/src/scanner.ts line 50: html_excerpt: z.string().default('') — no .max() constraint. The Python scanner caps at _HTML_EXCERPT_LIMIT = 6000 chars but the Zod schema accepts any length. If the Python cap is ever relaxed or another caller posts to the job, multi-megabyte blobs reach domain_events.metadata (jsonb column).

Fix

html_excerpt: z.string().max(8000).default(''),

[walidboulanouar]

Verified fixed — ScannerDomainSchema bounds html_excerpt at .max(8000) (packages/jobs/src/scanner.ts:69).