ACP

Record, Replay, and Test AI Agent Behavior

A local-first developer tool for recording, replaying, and testing AI agent behavior. Make agent execution deterministic, observable, and testable.

View on GitHubWatch DemoLearn More
By Hyper Grey / 2026
01 / What It Does

CORE WORKFLOW

No. 01

Trace Recording

Every agent step is captured with inputs, outputs, and state. Complete execution context in JSON format.

No. 02

Deterministic Replay

Same trace produces same replay behavior. Same final state after replay, every time.

No. 03

Step Inspector

Any step can be inspected. Inputs, outputs, and state visible at every decision point.

No. 04

Behavioral Testing

YAML-based test definitions. Tests catch logic regressions without depending on exact text.

No. 05

Trace Analysis

Tool highlights inefficiencies. Broken agent scenarios detected and explained.

02 / What This Proves

ALPHA VERSION CAPABILITIES

Core Capability

Deterministic Traces

Agent behavior can be recorded as a deterministic trace — every step is captured with inputs, outputs, and state.

01
Replay Engine

Exact Replay

That trace can be replayed exactly. Deterministic replay from recorded data.

02
Test Engine

Behavioral Testing

Developers can inspect and test agent behavior, not just outputs. Full step inspection and behavioral testing.

03
Visual Tools

VS Code Extension

Traces View, Steps View, and Trace Inspector Panel for visual step-by-step inspection.

04
Command Line

CLI Tools

npm run inspect, replay, test, and analyze commands. Interactive inspection and custom YAML tests.

05
Debugging

Failure Analysis

Tool explains why agents break. Detects high step count, memory growth, repeated calls, and more.

06
03 / Success Metrics

WHAT WE PROVE

01 / DETERMINISTIC

DETERMINISTIC REPLAY

Same trace produces same replay behavior. Same final state after replay. Every time, guaranteed.

02 / FULL

FULL INSPECTION

Any step can be inspected. Inputs, outputs, and state visible at every decision point.

03 / BEHAVIORAL

BEHAVIORAL TESTS

Tests catch logic regressions. Tests don't depend on exact text. YAML-based definitions.

04 / FAILURE

FAILURE ANALYSIS

Tool highlights inefficiencies. Broken agent scenario exists. Tool explains why it broke.

04 / Get Started

QUICK START

1. Run Agent

  • npm start
  • or --broken

2. Inspect

  • npm run inspect
  • traces/file.json

3. Test

  • npm run test
  • tests/basic.yaml

4. Analyze

  • npm run analyze
  • traces/file.json

Links

License

  • MIT
  • Open Source

© 2026 Hyper Grey. All rights reserved.

Agent Control Plane — Built with precision.