Fabraix Red Team Playground
Free · Open source
AI security researchers
PassControl
FreemiumAPISelf-Hosted
Every AI agent in your fleet sharing the same provider key means a compromised agent, a runaway spend loop, or a policy violation takes down everything — or forces a full key rotation that breaks every other agent at once. PassControl exists to end that shared-key problem.
Similar tools
Fabraix Red Team PlaygroundFreeFabraix Red Team Playground is an open-source, MIT-licensed testing environment structured…
EclusiaFreemiumEclusia runs locally on your machine and replaces names, IBANs, phone numbers, API keys, and…
Patronus ScannerFreemiumPatronus AI positions itself as evaluation and monitoring infrastructure for production LLM…
ArgotFreeArgot runs statistical checks against your repository's own commit history to flag code that…
Summary
PassControl sits between your agents and your LLM provider, giving each agent its own scoped credential while keeping the actual provider key server-side only. Before a call reaches OpenAI or any other provider, PassControl checks the agent's identity, its kill state, the model scope it's allowed, and whether its budget is still available — then returns a signed receipt so you have a verifiable record. Suspend one agent and the rest of the fleet keeps running. The product is in private beta for the hosted cloud path, which means production teams with strict uptime requirements should treat it as early-stage. Self-hosting via npm is available, but the vendor is explicit that the system has not yet been independently audited.
Bottom line: PassControl fits a team running multiple agents against a shared provider key who needs to suspend one bad actor without touching the rest — it does not yet fit a team that requires an audited security control before their legal or compliance team will sign off.
Find your fit
Before you choose PassControl
More options in Guardrails & Safety. Compare pricing, platforms, and features side by side.
Eclusia
Freemium
Legal professionals handling confidential cases
Pros
- Provider keys stay server-side and never enter agent environments, so a compromised agent environment cannot leak credentials that affect your entire provider account.
- Per-agent suspension without key rotation, which means taking a misbehaving agent offline does not require coordinating a credential change across every other agent in the fleet.
- Per-agent budget and rate-limit enforcement, so a spend spike from one agent hits a ceiling rather than accumulating across your entire billing account before you notice.
- Signed receipts on every allowed call, giving operators an auditable record of which agent called which model and when — without needing to reconstruct that from provider logs.
- Self-hosted deployment path available via npm with source-available code under BSL 1.1, so teams with infrastructure control requirements are not forced onto the cloud path.
Cons
- The hosted cloud is invite-only beta, which means teams that cannot run their own infrastructure and need a production-ready gateway today have no path forward until access opens — at that point they evaluate alternatives.
- The vendor explicitly states the system has not yet been independently audited. Any team whose security review, compliance posture, or legal sign-off requires a third-party audit report cannot deploy PassControl as a security control until that audit exists.
- The live demo simulates the model response and does not contact a real provider, so you cannot validate latency impact on actual provider calls or observe real-world receipt behavior before committing to integration.
- Teams who outgrow the scoped-credential identity model and need stronger cryptographic agent identity must implement Passport authentication separately — the docs describe it as available, but it is an additional integration step, not the default path.
About
- API Available
- Yes
- Self-Hosted
- Yes
- Last Updated
- August 16, 2026
Best For
Who it's for
- Teams running multiple AI agents
- Operators needing granular control
- Environments requiring key isolation
What it does well
- Secure agent access to LLM providers
- Per-agent budget and rate-limit enforcement
- Selective suspension without key rotation
- Audit trail via signed receipts
Frequently Asked Questions
- Is PassControl free?
- PassControl offers a free option alongside paid upgrades. Check the current allowances and terms before choosing a plan.
- Does PassControl have an API?
- Yes. PassControl exposes a developer API. See the official documentation at https://passcontrol.vertias.eu for details.
- Can I self-host PassControl?
- Yes. PassControl supports self-hosting on your own infrastructure.
Curated lists that include this category
Shared provider keys create fleet-wide risk
Every AI agent in your fleet sharing the same provider key means a compromised agent, a runaway spend loop, or a policy violation takes down everything — or forces a full key rotation that breaks every other agent at once.
How it works
PassControl sits between your agents and your LLM provider, giving each agent its own scoped credential while keeping the actual provider key server-side only. Before a call reaches OpenAI or any other provider, PassControl checks the agent’s identity, its kill state, the model scope it’s allowed, and whether its budget is still available — then returns a signed receipt so you have a verifiable record. Suspend one agent and the rest of the fleet keeps running. It offers an API and a self-hosted option.
Use cases
Secure agent access to LLM providers. Per-agent budget and rate-limit enforcement. Selective suspension without key rotation. Audit trail via signed receipts.
Pros
Provider keys stay server-side and never enter agent environments, so a compromised agent environment cannot leak credentials that affect your entire provider account. Per-agent suspension without key rotation. Per-agent budget and rate-limit enforcement. Signed receipts on every allowed call.
Limitations
The hosted cloud is invite-only beta, which means production teams with strict uptime requirements should treat it accordingly. The vendor states the system has not yet been independently audited. The live demo simulates the model response and does not contact a real provider.
Who it is for / who should skip it
Best for teams running multiple AI agents, operators needing granular control, and environments requiring key isolation. Teams that cannot run their own infrastructure or whose security review requires a third-party audit should skip it until access opens and an audit exists.
Contribute
- This listing has been claimed by the PassControl team.
- Embed a badge for your README or site.
Sign in to rate this tool or suggest an edit.