What is the HUMMBL AI Governance Arbiter audit?

The HUMMBL Arbiter audit is an evidence-backed A-F scoring system for AI agent codebases. It runs automated checks covering Bandit security analysis, test coverage, stdlib-only compliance, and governance primitive adoption to produce a reproducible grade.

How does the HUMMBL code quality audit work?

The audit works in three steps: submit your repository URL, HUMMBL Arbiter runs automated checks across security, test coverage, dependency hygiene, and governance primitives, then you receive a scored report with letter grade and remediation guidance — no agents to install and no config files to write.

What does the HUMMBL audit score mean?

Scores map to letter grades: A (90-100) indicates production-ready governance posture, B (80-89) shows minor gaps, C (70-79) indicates moderate risk, D (60-69) shows significant gaps, and F (below 60) indicates critical issues. All HUMMBL open-source repositories score A (95+).

Code Quality Audits

Evidence-backed A-F scoring for your AI agent codebase. Know exactly where you stand before your next deployment, compliance review, or funding round.

How It Works

Three steps from repo to report. No agents to install, no config files to write.

Submit Your Repo

Send us a link to your public or private repository. We support GitHub, GitLab, and Bitbucket. Private repos are handled under NDA.

Arbiter Scans

Our Arbiter engine runs 5 analysis passes: lint quality, security vulnerabilities, cyclomatic complexity, dead code detection, and duplication analysis.

Receive Your Report

Get a branded, shareable HTML report with an overall A-F grade, per-category breakdowns, and actionable remediation steps.

What We Measure

Five dimensions of code health, each weighted and scored independently.

Lint Quality

Style consistency, formatting violations, naming conventions, and static analysis warnings across your entire codebase.

Security

Bandit and Semgrep scans for known vulnerability patterns, hardcoded secrets, injection risks, and insecure defaults.

Complexity

Cyclomatic complexity per function and module. High complexity correlates with bugs, test gaps, and onboarding friction.

Dead Code

Unreachable functions, unused imports, orphaned modules, and stale test fixtures that inflate your maintenance burden.

Duplication

Copy-paste detection across files. Duplicated logic is a leading indicator of future divergence bugs and wasted effort.

Grade Scale

Scores map to letter grades with clear thresholds. No subjective judgment -- pure evidence.

90 - 100 Excellent

80 - 89 Good

70 - 79 Adequate

60 - 69 Needs Work

0 - 59 Critical

Public Audits

Open-source repositories we have audited. Click a report to see the full breakdown.

Repository	Grade	Score	Date	Report
hummbl-dev/base120	A	100.0	2026-04-14	View report
hummbl-dev/hummbl-governance	A	99.5	2026-04-14	View report
hummbl-dev/arbiter	A	98.0	2026-03-25	View report
hummbl-dev/mcp-server	A	100.0	2026-03-25	View report
hummbl-dev/agentic-patterns	A	100.0	2026-03-25	View report

Get your free audit

Send us your repo and receive a full Arbiter report within 48 hours. No cost, no commitment.

Request Free Audit Book a Call