Item detail

LiteLLM-Labs/litellm-agent-control-plane

RepoRadar surfaced LiteLLM-Labs/litellm-agent-control-plane — a unified agent control plane for — into the MIT unified agent control plane from LiteLLM tha section, where it sits at Gold tier with a 'try now' verdict. Its strongest signal is workflow potential, scored 9.9 out of 10.

Score8.4
Popularity1009.0
Risklow
TierGold
Score breakdown
Usefulness9.0
Novelty8.0
Momentum8.0
Maturity9.1
Open-source/build8.4
Evidence7.2
Workflow potential9.9
Setup ease8.8

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

Why it matters

Useful for engineering teams running multiple AI agent runtimes side by side (OpenCode, Hermes, Claude Managed Agents, Cursor Agents API, Deep Agents) who want one API + one UI + one set of credentials instead of a per-runtime console: LiteLLM Agent Control Plane is the MIT unified control plane that sits on top of any runtime and exposes one API to create / run / schedule / audit agents; for orga

Who should use it

Engineering teams running multiple AI agent runtimes side by side (OpenCode, Hermes, Claude Managed Agents, Cursor Agents API, Deep Agents) who want one API + one UI + one set of credentials instead of a per-runtime consoleOrganizations that adopted Claude Managed Agents and want to bring their own runtime (OpenCode, Hermes, Deep Agents) without rewriting the upstream code that drives the agent (the unified API shape means the same `create agent` / `run agent` / `list sessions` call works against any registered runtime)Teams that need persistent agent sessions across runs (the control plane stores session state in Postgres, so a long-horizon agent can resume after a restart without losing context)Product teams that want agents on a CRON schedule (the schedule surface stores cron expressions + agent IDs and triggers runs at the scheduled time)Engineering teams that want agents to remember context across sessions (the memory surface stores per-agent memory that the next session's prompt can reference)Security teams that want one audit log of every agent run across every runtime instead of a per-runtime audit trail (the Postgres schema is uniform)Adopters who want one-line local install (`docker compose --profile opencode up` starts LAP + Postgres + the OpenCode runtime + auto-registers the runtime in the UI)Teams that want a typed Rust core for the runtime (Cargo.lock + Cargo.toml + build.rs + Dockerfile + compose.yaml — production-grade engineering)Organizations that need access control so developers can create and run agents without direct Bedrock / Anthropic console access (the control plane mediates every API call)Users evaluating different agent runtimes who want to swap OpenCode for Hermes (or stack multiple runtimes) without changing the upstream code (the unified API shape is the abstraction)

Who should skip it

Move on from LiteLLM-Labs/litellm-agent-control-plane if the licensing terms, language support, or platform requirements do not fit your project.

About this signal

LiteLLM-Labs/litellm-agent-control-plane is tracked by RepoRadar as a unified agent control plane for in the MIT unified agent control plane from LiteLLM tha section. It was first seen on 2026-06-25 and last updated on 2026-06-25. The current verdict is 'try now' with a Gold tier and easy setup difficulty. The standout signals for LiteLLM-Labs/litellm-agent-control-plane are workflow potential (9.9) and maturity (9.1), while evidence quality (7.2) trails — that balance shapes where it fits best. This page summarizes the evidence RepoRadar has captured from captured source metadata. The score, tier, risk label, and verdict on this page are never influenced by sponsorship, ads, or tips — they reflect only the usefulness, popularity, novelty, momentum, maturity, and evidence signals described in the RepoRadar methodology.

How this item is evaluated

RepoRadar assigned LiteLLM-Labs/litellm-agent-control-plane a composite score of 8.4 out of 10, placing it in the Gold tier. This score combines weighted sub-signals: usefulness (35%), novelty (18%), momentum (14%), maturity (10%), open-source/build quality (7%), evidence quality (6%), workflow potential (6%), and setup ease (4%). Popularity is tracked separately at 1009.0 and never affects the composite score or tier. The risk label of 'low' reflects inherent user-impacting hazards, not generic novelty. Items with no risk flag may still require normal code review before production use.

Risk explanation

**One-line install default master key is `sk-local`; change it before any non-local deployment.** The README is explicit that `sk-local` is the default master key and is for local install only. Adopters running LAP on a shared host, in CI, or in production must change the master key to a strong random value before exposing the web UI / API, and should rotate the master key on the same cadence as any other production credential. The web UI signs in with the master key, so a leaked key is full control of the agent fleet; **Provider credentials are configured in Settings after install; configure before running agents against hosted model providers.** The README is explicit that adopters must add provider credentials in Settings before running agents against a hosted model provider (OpenAI, Anthropic, etc.). LAP mediates every API call to the underlying model provider, so the credentials live in the LAP config + the Postgres audit log — protect the Postgres credentials with the same care as the provider credentials themselves; **Runtime profiles register after health checks pass; verify the runtime is healthy before scheduling agent runs on it.** Stacked-profile deploys (`--profile opencode --profile deepagents`) register each runtime in the LAP UI after its health check passes. Adopters should verify the runtime is actually healthy in the UI before scheduling agent runs on it — a runtime that fails its health check will not be registered, and an agent scheduled against a non-registered runtime will fail at run time.

Evidence links

Closest alternatives / related signals

litellmlitellm-agent-control-planelaplite-llmagent-control-planecontrol-planeagent-runtimeopencode