ToolExperimental

agentplane

Name: agentplane
Rating: 3.0 (72 reviews)
Author: basilisk-labs

by basilisk-labs

Governed CLI for deterministic, auditable agent workflows

TypeScript

Updated Jul 25, 2026

Stars

Forks

View on GitHub

Overview

Provides a governed CLI framework for building deterministic, auditable AI agent workflows. Uses policy-driven configuration, deterministic execution traces, and CLI-driven orchestration to make agent runs reproducible and inspectable. Includes features for strict logging, step-by-step replay, and policy enforcement to support compliance and post-hoc review. The framework supports auditable AI agent workflows and governance to help teams verify behavior and maintain an agent track record.

Need implementation help?

Expert guidance available

Key Benefits

As agents become more autonomous, being able to reproduce and audit their decision paths is essential for trust and liability. Multi-agent trust and auditable runs come from governance and determinism to agent workflows so teams can verify behavior, investigate failures, and maintain an agent track record. This matters for multi-agent trust because auditable runs and policy checks turn opaque interactions into evidence you can evaluate and compare.

Ideal For

Teams building developer-facing agent pipelines that need reproducible runs, policy enforcement, and post-run auditability. For reproducible runs, consider patterns like LLM-based evaluation patterns to ensure consistent outcomes, while policy enforcement helps codify guardrails, and post-run auditability provides a traceable record for reviews.

Real-World Examples

Run and replay agent workflows to reproduce behaviors during incident investigation
Enforce policies and guardrails at runtime for compliance-sensitive agent tasks
Record deterministic traces and logs to build an agent track record for evaluations