
Founder & CEO 🐕@cerbereag The control infrastructure for autonomous AI agents operating across real-world @luviaplace to be passionate about what I've done
Joined September 2026
- Tweets383
- Following280
- Followers107
- Likes556
You put a cohort of agents in prod. Because you want fast production.
But then Agent X deletes Agent Y's work.
And Agent Z just loops. Loops. Loops.
In the end, you move nowhere.
A successful project needs coordination.
Frameworks like LangChain or CrewAI let agents work together. Let them talk.
Cerbere Cohort does the rest:
• Isolated workspace for each agent
• Conflict-aware commits
• Deadlock detection
• Verifiable history
If Cerbere AG stops your agent from doing anything stupid,
Cohort stops agents from fighting each other.
In the end, you have your army of agents running 24/7.
And your project finishes on time. Or even early.
Let's build it together.
cerbereag.site
github.com/chrismsmr-celcom/…
github.com/chrismsmr-celcom/…
4
71
Autonomous AI doesn't need more intelligence. It needs guardrails.
Since August, I've been building Cerbere:
Cerbere AG - stops your agent from doing something stupid in prod. It blocks the bad tool call before it runs.
Cerbere Cohort - stops your agents from stepping on each other. No conflicting writes, no corrupted shared state.
One protects how an agent acts.
The other protects how agents work together.
Open source. Free. 1000+ commits in.
Built from Kinshasa.
cerbereag.site
github.com/chrismsmr-celcom/…
github.com/chrismsmr-celcom/…
3
51
I'm building Cerbere Cohort, an open-source coordination layer for multi-agent systems.
Frameworks like LangGraph, CrewAI and AutoGen help agents work together.
Cerbere Cohort focuses on a different problem: keeping their shared state consistent.
Isolated workspaces
Conflict-aware commits
Resource leases
Deadlock detection
Verifiable event history
We're not building another orchestration framework. We're building a layer that complements existing ones.
The project is still early, and I'd love to have developers involved.
3
4
73
Most secure the prompt.
No one secures what the agent DOES after.
Cerbrere does both.
Prompt injection ✅
Tool call protection ✅
Taint tracking ✅
Signed decisions ✅
Self-hosted ✅
Observability ✅
MCP security ✅
Humain approval✅
Your data never leaves your servers.
We don't filter prompts. We control actions.
For the next 6 months, Cerbrere is free and open source.
DMs are open.
Who is Cerbere AG for?
Not for everyone.
For:
• Companies securing their AI agents in production
• Teams that need SOC 2 / GDPR proof
• Devs who want to control LLM costs, not just watch them explode
• Any org running autonomous agents
If your agent can read, write, delete and send, you need a guardian.
Not a dashboard.
3
7
207
Cerbere is self-hosted and settles in a few minutes.
To protect you, we do not retrieve any of your data. Sending them to the cloud would be just moving the attack surface elsewhere.
So we built Cerbere to run directly in your infrastructure. Your safety remains at home. Your privacy too.
cerbereag.site/docs
Open source:
github.com/chrismsmr-celcom/…
1
6
90
an AI agent goes into a loop on a Sunday night.
Result: 890M tokens, a $2,450 bill, and $7.47 per request instead of $0.0075.
A cost spike is rarely a billing bug. More often, it’s the first sign of a compromised or runaway agent.
That’s what we monitor at Cerbere.
We didn't make 3 plans randomly.
We made 3 realities.
Team at $599/mo is the one we built for startups going to production.
Why?
1. You are not alone anymore
5 seats. Everyone sees what your agents do, in real time.
2. Volume changes everything
500k protected actions / mo
25k LLM analyses included
Enough for prod. Not paying for empty volume.
3. In prod, you need proof
Signed audit export + Priority Slack
When someone asks "prove your agents are controlled", you export. Done.
Enterprise is for banks and EU AI Act. Not for now.
Team is what we would recommend to a founder friend asking "I'm going to prod, what do I take?"
That’s why we recommend it!
Split the .env secret into pieces. Encode it. Send it out through an http_post. Still blocked.
This is yesterday's attack, seen from the Cerbere dashboard. The ops-agent reads .env, and then an injection tries to sneak the secret to an attacker URL. Taint tracking catches it before the call runs.
Click the decision to open the trace: the span timeline, which policies fired (sensitive taint reaching a high-impact sink), and the exact payload it tried to send.
Open source 👇
github.com/chrismsmr-celcom/…
Cerbere-AG is a three-layer security guardrail for AI agents: Regex, ML, and an LLM Judge.
AI agents are unpredictable. They improvise and act with full autonomy. I built Cerbere-AG (named after the three-headed dog, Cerberus) so that autonomy doesn’t come at the cost of control.
Why three layers? Our benchmarks show it:
→ Regex alone lets more than 30% of attacks through.
→ ML pushes recall up to 100% in some of our tests, but raises false positives.
→ The LLM Judge resolves the ambiguous cases ML can’t classify.
It also looks at context. Cerbere-AG follows tool calls, checks whether they’re legitimate, and then allows them, blocks them, or flags them for real-time human approval. Agents keep their autonomy, and humans keep control.
Everything is audited too. The collector gathers spans and traces into an audit trail that is independent from the agent and supports compliance.
And like any tool, it has limits. That’s why observability matters: security isn’t only about blocking attacks, it’s about understanding what happened, what was allowed, what was blocked, and why.
Open source and free for now. Try it and give me your honest feedback 👇
github.com/chrismsmr-celcom/…
app.cerbereag.site
6
172






