agent-governance-toolkit vs agentic_security
agent-governance-toolkit is much bigger: 6.2k stars against 2.0k. Over the days we have tracked them agent-governance-toolkit moved +51.2% and agentic_security +4.7%, so agent-governance-toolkit is growing faster right now.
Neither one leads on the six capability axes, so the choice comes down to which of them fits the way you already work.
Stars and commit dates come from our own daily tracking. The six axes are read off each project's documentation by our review pipeline, so they describe what a project says about itself, not what we measured in its code.
Where they stand today
Deterministic, application-level enforcement that intercepts every tool call and makes denied actions structurally impossible (fail-closed governance plus tamper-evident audit), rather than relying on prompt-level controls.
- Stars
- 6.2k
- Tracked growth
- +51.2%
- Maturity
- ●●●●●
- Last commit
- 1d ago
- Language
- Python
- License
- MIT
- Cost to run
- Free, open-source
Agentic Security focuses on agent/workflow-focused, multimodal (text, image, audio) and multi-step attack/fuzzing capabilities (including RL-based and dataset mutation approaches) to evaluate LLM vulnerabilities beyond single-shot prompt tests.
- Stars
- 2.0k
- Tracked growth
- +4.7%
- Maturity
- ●●●●●
- Last commit
- 3d ago
- Language
- Python
- License
- Apache-2.0
- Cost to run
- Your LLM API key, usage billed by the provider
Six axes, head to head
Each axis runs 0 to 5. The label under a score is what that project's own docs claim, not a category average.
| Axis | agent-governance-toolkit | agentic_security |
|---|---|---|
Context depth How much of your codebase it sees before it answers: the open diff, the diff plus related files, or the whole repository. | ●●●●● Action-level only | ●●●●● Prompt-only |
Noise control How it keeps output volume down — severity thresholds, deduplication, incremental runs over new commits only. | ●●●●● Approval gating | ●●●●● Thresholds/config |
Customization How far it bends to your team: custom rules, prompts, style guides, per-path config. | ●●●●● YAML & APIs | ●●●●● Config + modules |
Privacy Whether your code stays on your own infrastructure: fully local, self-hostable, or cloud API only. | ●●●●● Fully local runnable | ●●●●● Self-hostable |
Model freedom Whether you can point it at any provider, or it is wired to one. | ●●●●● Bring-your-own key | ●●●●● Bring-your-own-key |
Setup ease What it takes to get a first useful run out of it. | ●●●●● One-command start | ●●●●● One-command start |
Which one to pick
Pick agent-governance-toolkit if…
Deterministic enforcement — pick AGT when you need fail-closed, application-level policy enforcement and tamper-evident auditing that prevents agents from performing denied actions.
Pick agentic_security if…
Easy setup — pip-installable and runnable locally with configurable datasets and modules to quickly run multimodal, multi-step vulnerability scans against your LLM endpoints.
What people want from each one
microsoft/agent-governance-toolkit
Hacker News: Agent Governance Toolkit drew 5 points and 0 comments.
msoedov/agentic_security
Hacker News: Agentic Security – LLM Vulnerability Scanner drew 2 points and 0 comments.
Questions people ask
Is agent-governance-toolkit better than agentic_security?
Neither one leads on the six capability axes, so the choice comes down to which of them fits the way you already work. agent-governance-toolkit is worth picking when deterministic enforcement — pick AGT when you need fail-closed, application-level policy enforcement and tamper-evident auditing that prevents agents from performing denied actions.
Which of agent-governance-toolkit and agentic_security keeps my code private?
agent-governance-toolkit: Fully local runnable (5/5). agentic_security: Self-hostable (4/5).
What does each one cost to run?
agent-governance-toolkit: Free, open-source. agentic_security: Your LLM API key, usage billed by the provider.
Full profiles: microsoft/agent-governance-toolkit and msoedov/agentic_security. Everything else in Security & guardrails.