Bounty Engineer Agent Commerce
io.github.gatoprd/bounty-engineer · v1.3.0 · score 89/100 · checked 2026-11-22
Automated, read-only signal from public data, not an audit or an accusation. Maintainers can dispute any finding. How grading works
Remote MCP storefront for agent routing, verification, web extraction and x402-paid tools.
Source: https://github.com/gatoprd/InstaVeil
Findings
| Severity | Rule | Detail |
|---|---|---|
| Low | MW005 | Remote endpoint lists 22 tool(s) to anyone, no credentials needed. |
| Medium | MW102 | 1 tool(s) can execute commands or delete data (execute_paid_product). Require confirmation before the agent calls them. |
What was checked
- https://bounty-engineer-x402-prod.onrender.com/mcp: answers without auth
Public tools: list_products, get_commerce_radar_offer, quote_batch_web_intelligence, list_premium_agent_packs, qualify_premium_buyer, quote_premium_workload, quote_premium_delivery, recommend_paid_tool, get_purchase_instructions, quote_autonomous_purchase, get_product_proof, quote_instant_work, paid_web_extract, paid_funding_loop, paid_funding_rates, paid_json_contract_check, paid_x402_payment_preflight, paid_mcp_manifest_security, paid_merchant_feed_audit, paid_agent_purchase_firewall, execute_paid_product, get_workflow_recipe
Maintainer? Add the badge
Grades refresh daily. Think a finding is wrong? Dispute this grade.
[](https://www.mcpwatch.io/s/io-github-gatoprd-bounty-engineer-c7d4a8a2)