Scanner deep-dive
Kimi K3 by Moonshot AI ↗
General-Purpose LLM · agentic-v1 · scored on 66/66 Python repositories. Strict scoring (unfinished repos counted as misses).
59.4
F3 (strict)
60.7
F2 (strict)
58.2%
Recall (strict)
73.0%
Precision
66/66
Repos scored
kimi-k3
Model
$21
Total cost
380s
Avg latency
§
Per-repository breakdown
Each bar shows true positives, false positives, and misses on one repository; bar length is proportional to that repo's labeled vulnerabilities. Ranked by F2.
| Repository | TP | FP | FN | Recall % | F2 |
|---|---|---|---|---|---|
| pythonssti | 2 | 0 | 0 | 100.0 | 100.0 |
| vfapi | 9 | 0 | 0 | 100.0 | 100.0 |
| vampi | 13 | 2 | 2 | 86.7 | 86.7 |
| dsvw | 22 | 4 | 5 | 81.5 | 82.1 |
| python-app | 16 | 2 | 5 | 76.2 | 78.4 |
| vulnpy | 61 | 10 | 19 | 76.2 | 78.0 |
| insecure-web | 7 | 2 | 2 | 77.8 | 77.8 |
| vc-kimi-code-seeded-v2-legal-case-django | 20 | 8 | 6 | 76.9 | 75.8 |
| vc-codex-high-seeded-v2-hr-payroll-django | 19 | 7 | 6 | 76.0 | 75.4 |
| vulnerable-api | 10 | 1 | 4 | 71.4 | 74.6 |
| vc-kimi-code-seeded-v2-hr-payroll-django | 21 | 5 | 8 | 72.4 | 73.9 |
| dsvpwa | 23 | 5 | 9 | 71.9 | 73.7 |
| vc-codex-high-seeded-v2-education-lms-django | 19 | 10 | 6 | 76.0 | 73.6 |
| vulnerable-tornado-app | 10 | 3 | 4 | 71.4 | 72.5 |
| vc-kimi-code-seeded-v2-property-management-fastapi | 21 | 5 | 9 | 70.0 | 71.9 |
| vc-kimi-code-seeded-v2-support-desk-fastapi | 20 | 7 | 8 | 71.4 | 71.9 |
| owasp-web-playground | 19 | 3 | 9 | 67.9 | 70.9 |
| dvblab | 15 | 4 | 7 | 68.2 | 70.1 |
| vc-codex-high-seeded-v2-marketplace-commerce-fastapi | 17 | 5 | 8 | 68.0 | 69.7 |
| vc-claude-code-seeded-v2-legal-case-django | 21 | 6 | 10 | 67.7 | 69.5 |
| vc-codex-high-seeded-v2-crm-saas-django | 17 | 6 | 8 | 68.0 | 69.1 |
| vc-codex-seeded-v2-logistics-dispatch-fastapi | 19 | 4 | 11 | 63.3 | 66.4 |
| vc-kimi-code-seeded-v2-marketplace-commerce-fastapi | 18 | 10 | 9 | 66.7 | 66.2 |
| vc-kimi-code-seeded-v2-fintech-lending-fastapi | 21 | 7 | 12 | 63.6 | 65.6 |
| vc-kimi-code-seeded-v2-logistics-dispatch-fastapi | 20 | 9 | 11 | 64.5 | 65.4 |
| vc-codex-seeded-v2-marketplace-commerce-fastapi | 18 | 5 | 11 | 62.1 | 64.7 |
| vc-codex-high-seeded-v2-fintech-lending-fastapi | 20 | 19 | 9 | 69.0 | 64.5 |
| vc-kimi-code-seeded-v2-healthcare-clinic-django | 20 | 15 | 10 | 66.7 | 64.5 |
| threatbyte | 16 | 5 | 10 | 61.5 | 64.0 |
| pygoat | 47 | 9 | 31 | 60.3 | 63.9 |
| vc-claude-code-seeded-v2-fintech-lending-fastapi | 18 | 7 | 11 | 62.1 | 63.8 |
| vc-codex-high-seeded-v2-logistics-dispatch-fastapi | 18 | 7 | 11 | 62.1 | 63.8 |
| vc-claude-code-seeded-v2-support-desk-fastapi | 20 | 4 | 14 | 58.8 | 62.5 |
| extremely-vulnerable-flask-app | 18 | 2 | 14 | 56.2 | 60.8 |
| intentionally-vulnerable-python-application | 4 | 1 | 3 | 57.1 | 60.6 |
| vc-codex-high-seeded-v2-support-desk-fastapi | 16 | 5 | 12 | 57.1 | 60.2 |
| vc-claude-code-seeded-v2-logistics-dispatch-fastapi | 19 | 7 | 14 | 57.6 | 60.1 |
| vc-codex-seeded-v2-support-desk-fastapi | 17 | 5 | 13 | 56.7 | 59.9 |
| vc-codex-high-seeded-v2-property-management-fastapi | 16 | 14 | 10 | 61.5 | 59.7 |
| vulnerable-flask-app | 12 | 5 | 9 | 57.1 | 59.4 |
| vc-claude-code-seeded-v2-crm-saas-django | 16 | 7 | 12 | 57.1 | 59.3 |
| vc-claude-code-seeded-v2-education-lms-django | 18 | 6 | 14 | 56.2 | 59.2 |
| vulnerable-python-apps | 12 | 3 | 10 | 54.5 | 58.3 |
| vc-claude-code-seeded-v2-hr-payroll-django | 15 | 6 | 12 | 55.6 | 58.1 |
| vc-codex-seeded-v2-hr-payroll-django | 21 | 4 | 18 | 53.8 | 58.0 |
| vc-codex-seeded-v2-property-management-fastapi | 17 | 7 | 14 | 54.8 | 57.4 |
| vc-claude-code-seeded-v2-healthcare-clinic-django | 16 | 8 | 13 | 55.2 | 57.1 |
| vc-codex-seeded-v2-crm-saas-django | 18 | 5 | 16 | 52.9 | 56.6 |
| vc-kimi-code-seeded-v2-education-lms-django | 16 | 14 | 12 | 57.1 | 56.3 |
| python-insecure-app | 4 | 0 | 4 | 50.0 | 55.6 |
| vc-claude-code-seeded-v2-property-management-fastapi | 17 | 4 | 16 | 51.5 | 55.6 |
| vc-codex-seeded-v2-legal-case-django | 17 | 5 | 16 | 51.5 | 55.2 |
| vc-codex-high-seeded-v2-healthcare-clinic-django | 14 | 10 | 12 | 53.8 | 54.7 |
| vc-codex-seeded-v2-education-lms-django | 18 | 9 | 17 | 51.4 | 53.9 |
| vc-claude-code-seeded-v2-marketplace-commerce-fastapi | 16 | 6 | 16 | 50.0 | 53.3 |
| vc-codex-seeded-v2-healthcare-clinic-django | 21 | 12 | 20 | 51.2 | 53.3 |
| lets-be-bad-guys | 12 | 5 | 12 | 50.0 | 53.1 |
| vc-kimi-code-seeded-v2-crm-saas-django | 14 | 11 | 13 | 51.8 | 52.6 |
| djangoat | 25 | 9 | 27 | 48.1 | 51.7 |
| damn-vulnerable-graphql-application | 16 | 6 | 20 | 44.4 | 48.2 |
| damn-vulnerable-flask-application | 6 | 6 | 8 | 42.9 | 44.1 |
| flask-xss | 11 | 1 | 19 | 36.7 | 41.7 |
| vc-codex-seeded-v2-fintech-lending-fastapi | 14 | 6 | 23 | 37.8 | 41.7 |
| dvpwa | 8 | 5 | 15 | 34.8 | 38.1 |
| vc-codex-high-seeded-v2-legal-case-django | 6 | 19 | 19 | 24.0 | 24.0 |
| vulpy | 0 | 0 | 57 | 0.0 | 0.0 |
§
Detection by severity
| Severity | TP | FP | FN | Recall % |
|---|---|---|---|---|
| Critical | 142 | 0 | 12 | 92.2 |
| High | 468 | 2 | 284 | 62.2 |
| Medium | 451 | 1 | 459 | 49.6 |
| Low | 46 | 0 | 40 | 53.5 |
§
Detection by vulnerability class
| CWE family | TP | FP | FN | Recall % |
|---|---|---|---|---|
| Open Redirect | 40 | 0 | 0 | 100.0 |
| HTTP Header Injection | 2 | 0 | 0 | 100.0 |
| XPath Injection | 4 | 0 | 0 | 100.0 |
| Command / OS Injection | 47 | 0 | 2 | 95.9 |
| Insecure Deserialization | 37 | 0 | 3 | 92.5 |
| XML External Entities | 35 | 0 | 3 | 92.1 |
| Code Injection / RFI | 31 | 0 | 3 | 91.2 |
| SQL Injection | 71 | 0 | 7 | 91.0 |
| Path Traversal | 38 | 1 | 6 | 86.4 |
| Security Misconfiguration | 80 | 0 | 32 | 71.4 |
| Hardcoded Credentials | 48 | 0 | 20 | 70.6 |
| Server-Side Request Forgery | 28 | 2 | 13 | 68.3 |
| Denial of Service | 27 | 0 | 17 | 61.4 |
| Cross-Site Scripting | 67 | 0 | 43 | 60.9 |
| Sensitive Data Exposure | 81 | 0 | 74 | 52.3 |
| Missing Authentication / Authorization | 42 | 0 | 41 | 50.6 |
| Other | 406 | 0 | 471 | 46.3 |
| Broken Access Control / IDOR | 23 | 0 | 60 | 27.7 |
§
LLM operational metrics
53,273
Avg input tokens
10,894
Avg output tokens
385,393
Avg total tokens
380s
Avg latency / repo
1.5%
JSON repair rate
66
Total runs
§
Cost
$21
Total cost
$0.32
Cost / run
$0.016
Cost / 100 LOC
133,782
Lines of code scanned
66
Successful runs