realvuln v2.0
Dashboard Methodology Dataset Findings Roadmap GitHub ↗
Scanner deep-dive

Ornith 1.0 35B by OrnithAI ↗

General-Purpose LLM · agentic-v1 · scored on 58/66 repositories. Strict scoring (unfinished repos counted as misses).

Methodology note
Locally hosted; template fix. Open-weight GGUF (Q4_K_M) via Ollama through the standard agentic-v1 harness. Ornith's shipped GGUF template was replaced with the canonical Qwen3 template for tool calling. Completed 58/66. Cost $0.
23.3
F3 (strict)
24.9
F2 (strict)
22.0%
Recall (strict)
53.6%
Precision
58/66
Repos scored
ornith-q3
Model
Free
Total cost
243s
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.

True positiveFalse positiveMissed (FN)
pythonssti100 F2 · 100%
dsvw72 F2 · 70%
vulnerable-api62 F2 · 57%
vulnerable-flask-app60 F2 · 62%
vulnpy58 F2 · 60%
insecure-web56 F2 · 56%
vc-kimi-code-seeded-v2-healthcare-clinic-django55 F2 · 57%
vfapi53 F2 · 56%
vc-codex-high-seeded-v2-hr-payroll-django52 F2 · 52%
damn-vulnerable-flask-application52 F2 · 47%
vc-kimi-code-seeded-v2-crm-saas-django51 F2 · 48%
python-app50 F2 · 52%
vc-kimi-code-seeded-v2-education-lms-django49 F2 · 46%
vc-claude-code-seeded-v2-crm-saas-django48 F2 · 43%
vc-codex-high-seeded-v2-legal-case-django47 F2 · 44%
vc-kimi-code-seeded-v2-hr-payroll-django45 F2 · 45%
vc-codex-seeded-v2-hr-payroll-django44 F2 · 41%
vc-codex-seeded-v2-logistics-dispatch-fastapi43 F2 · 40%
dvpwa43 F2 · 39%
vc-codex-seeded-v2-fintech-lending-fastapi43 F2 · 41%
lets-be-bad-guys41 F2 · 38%
extremely-vulnerable-flask-app41 F2 · 38%
vc-codex-seeded-v2-support-desk-fastapi40 F2 · 40%
vc-codex-seeded-v2-legal-case-django39 F2 · 36%
vulnerable-tornado-app39 F2 · 36%
vc-claude-code-seeded-v2-education-lms-django38 F2 · 34%
vc-codex-high-seeded-v2-marketplace-commerce-fastapi36 F2 · 36%
dvblab36 F2 · 32%
vc-codex-seeded-v2-property-management-fastapi35 F2 · 32%
vampi34 F2 · 33%
pygoat33 F2 · 29%
vc-codex-seeded-v2-healthcare-clinic-django33 F2 · 29%
vc-claude-code-seeded-v2-logistics-dispatch-fastapi30 F2 · 27%
threatbyte25 F2 · 23%
vc-codex-seeded-v2-education-lms-django25 F2 · 23%
vc-claude-code-seeded-v2-hr-payroll-django25 F2 · 22%
djangoat0 F2 · 0%
dsvpwa0 F2 · 0%
flask-xss0 F2 · 0%
intentionally-vulnerable-python-application0 F2 · 0%
vulnerable-python-apps0 F2 · 0%
vc-claude-code-seeded-v2-fintech-lending-fastapi0 F2 · 0%
vc-claude-code-seeded-v2-legal-case-django0 F2 · 0%
vc-claude-code-seeded-v2-marketplace-commerce-fastapi0 F2 · 0%
vc-claude-code-seeded-v2-property-management-fastapi0 F2 · 0%
vc-claude-code-seeded-v2-support-desk-fastapi0 F2 · 0%
vc-codex-high-seeded-v2-crm-saas-django0 F2 · 0%
vc-codex-high-seeded-v2-education-lms-django0 F2 · 0%
vc-codex-high-seeded-v2-fintech-lending-fastapi0 F2 · 0%
vc-codex-high-seeded-v2-logistics-dispatch-fastapi0 F2 · 0%
vc-codex-high-seeded-v2-property-management-fastapi0 F2 · 0%
vc-codex-high-seeded-v2-support-desk-fastapi0 F2 · 0%
vc-codex-seeded-v2-crm-saas-django0 F2 · 0%
vc-kimi-code-seeded-v2-fintech-lending-fastapi0 F2 · 0%
vc-kimi-code-seeded-v2-legal-case-django0 F2 · 0%
vc-kimi-code-seeded-v2-marketplace-commerce-fastapi0 F2 · 0%
vc-kimi-code-seeded-v2-property-management-fastapi0 F2 · 0%
vc-kimi-code-seeded-v2-support-desk-fastapi0 F2 · 0%
RepositoryTPFPFNRecall %F2
pythonssti200100.0100.0
dsvw195870.472.0
vulnerable-api80657.162.5
vulnerable-flask-app1312861.959.6
vulnpy48493260.057.6
insecure-web54455.655.6
vc-kimi-code-seeded-v2-healthcare-clinic-django17171356.755.2
vfapi56455.653.2
vc-codex-high-seeded-v2-hr-payroll-django13111252.052.4
damn-vulnerable-flask-application71846.751.5
vc-kimi-code-seeded-v2-crm-saas-django1371448.150.8
python-app11141052.450.5
vc-kimi-code-seeded-v2-education-lms-django1371546.449.2
vc-claude-code-seeded-v2-crm-saas-django1221642.947.6
vc-codex-high-seeded-v2-legal-case-django1171444.046.6
vc-kimi-code-seeded-v2-hr-payroll-django13151644.845.1
vc-codex-seeded-v2-hr-payroll-django1682341.044.4
vc-codex-seeded-v2-logistics-dispatch-fastapi1271840.043.2
dvpwa941439.142.9
vc-codex-seeded-v2-fintech-lending-fastapi15132240.542.6
lets-be-bad-guys951537.540.9
extremely-vulnerable-flask-app1272037.540.8
vc-codex-seeded-v2-support-desk-fastapi12161840.040.5
vc-codex-seeded-v2-legal-case-django1292136.439.2
vulnerable-tornado-app53935.739.1
vc-claude-code-seeded-v2-education-lms-django1172134.437.7
vc-codex-high-seeded-v2-marketplace-commerce-fastapi9161636.036.0
dvblab731531.835.7
vc-codex-seeded-v2-property-management-fastapi10102132.334.7
vampi591033.333.8
pygoat2395529.533.4
vc-codex-seeded-v2-healthcare-clinic-django1252929.333.1
vc-claude-code-seeded-v2-logistics-dispatch-fastapi9112427.329.6
threatbyte682023.125.4
vc-codex-seeded-v2-education-lms-django8112722.925.2
vc-claude-code-seeded-v2-hr-payroll-django672122.224.8
djangoat00520.00.0
dsvpwa00320.00.0
flask-xss09300.00.0
intentionally-vulnerable-python-application0670.00.0
vulnerable-python-apps00220.00.0
vc-claude-code-seeded-v2-fintech-lending-fastapi00290.00.0
vc-claude-code-seeded-v2-legal-case-django00310.00.0
vc-claude-code-seeded-v2-marketplace-commerce-fastapi00320.00.0
vc-claude-code-seeded-v2-property-management-fastapi00330.00.0
vc-claude-code-seeded-v2-support-desk-fastapi00340.00.0
vc-codex-high-seeded-v2-crm-saas-django00250.00.0
vc-codex-high-seeded-v2-education-lms-django00250.00.0
vc-codex-high-seeded-v2-fintech-lending-fastapi00290.00.0
vc-codex-high-seeded-v2-logistics-dispatch-fastapi011290.00.0
vc-codex-high-seeded-v2-property-management-fastapi00260.00.0
vc-codex-high-seeded-v2-support-desk-fastapi00280.00.0
vc-codex-seeded-v2-crm-saas-django00340.00.0
vc-kimi-code-seeded-v2-fintech-lending-fastapi00330.00.0
vc-kimi-code-seeded-v2-legal-case-django00260.00.0
vc-kimi-code-seeded-v2-marketplace-commerce-fastapi00270.00.0
vc-kimi-code-seeded-v2-property-management-fastapi00300.00.0
vc-kimi-code-seeded-v2-support-desk-fastapi011280.00.0
§

Detection by severity

SeverityTPFPFNRecall %
Critical7106950.7
High179647927.2
Medium156763719.7
Low1205617.6
§

Detection by vulnerability class

CWE familyTPFPFNRecall %
HTTP Header Injection200100.0
XPath Injection400100.0
Insecure Deserialization2401266.7
XML External Entities2201264.7
Path Traversal2531464.1
SQL Injection3732757.8
Code Injection / RFI1701356.7
Command / OS Injection2302152.3
Open Redirect1702144.7
Server-Side Request Forgery1512240.5
Cross-Site Scripting3536335.7
Hardcoded Credentials1803534.0
Security Misconfiguration2207822.0
Sensitive Data Exposure24010518.6
Other117065515.2
Missing Authentication / Authorization935813.4
Denial of Service30377.5
Broken Access Control / IDOR40685.6
§

LLM operational metrics

307,619
Avg input tokens
7,913
Avg output tokens
315,531
Avg total tokens
243s
Avg latency / repo
45.5%
JSON repair rate
66
Total runs
§

Cost

Free
Total cost
117,112
Python LOC scanned
58
Successful runs

← Back to the leaderboard