Item detail

Floe-Labs/floe-guard

floe-guard is an MIT-licensed open-source budget guardrail for AI agents that hard-stops a runaway loop before it burns the bill, with a local, framework-agnostic Python core and an optional JS adapter. It is the first budget-guardrail project that ships as a library you can drop into any agent (LangChain, AutoGen, custom), and it is built by a small team that has been tracking agent-cost blow-ups

Score7.3
Popularity45.0
Risknone
TierSilver
Score breakdown
Usefulness8.0
Novelty8.0
Momentum6.0
Maturity6.3
Open-source/build8.4
Evidence7.2
Workflow potential8.8
Setup ease8.8

Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.

Why it matters

Useful for developers running agents in production who need a hard stop on runaway loops before the API bill spikes: install floe-guard, set a per-session token/USD budget, and let it cut the loop when the limit trips.

Who should use it

developers running agents in production on paid APIsLangChain / AutoGen / custom-agent users who want a hard stop on runaway loopsplatform teams that need a per-session budget cap with audit logs

Who should skip it

Skip if the source link, docs, or setup requirements do not match your workflow.

Risk explanation

No inherent user-impacting risk is flagged from the captured evidence.

Evidence links

Closest alternatives / related signals

agent-costguardrailbudgetlangchainautogenagent-safety