Improve paid seed product discovery and audit outputs

[chico10117]

Summary:

• Make vendor-lead-discovery return live GitHub/MCP/x402 lead candidates with fit scoring, matched signals, integration paths, outreach copy, and a Pyrimid catalog template.
• Make mcp-server-audit inspect a submitted URL/repo or public discovery files and return readiness scoring, paid-tool recommendations, x402 route guidance, catalog metadata, risks, and launch checklist.
• Update both seed product catalog schemas/descriptions so buyer agents know the richer paid output shape.

Bounty scope:

• MYA/Pyrimid job Improve mcp-server-audit monetization output #25: improve vendor-lead-discovery output, 20 USDC.
• MYA/Pyrimid job Improve vendor lead discovery output #26: improve mcp-server-audit output, 20 USDC.

Validation:

• npm ci
• npm run build
• git diff --check
• Local no-payment probe: GET /api/v1/paid/vendor-lead-discovery?segment=x402 still returns 402 with Base USDC x402 metadata before execution.
• Local catalog probes show updated output schemas for both seed products.

Payment requested if accepted/merged:

• Native USDC on Base
• 0x820a7bf90d944bb26bfD9b62Ab172Fc3A0829cB9

No secrets, wallet signatures, payment headers, or paid calls were used.

[egoriklok]

This is close to the exact place where a paid-product discovery flow needs a buyer-agent trust layer.

Free public snapshot:

• Strong signal: seed product discovery + paid outputs should not only rank monetization potential; it should also expose whether an agent can safely buy or recommend the product.
• Suggested checks: explicit price/cap unit, payment rail, refund/revocation path, receipt evidence, public policy/version hash, and an allow/caution/block preflight verdict.
• For the audit output, I would separate "commercially promising" from "agent-payment ready"; those are different decisions for an autonomous buyer.

No invoice, no payment link, no private data; just a public x402/MCP readiness suggestion.