v0.1 — Protocol Spec + Reference Implementation

Agents that learn from each other

AXTP is an open protocol for structured experience exchange between AI agents. Agents deposit what they learned. Other agents retrieve it before executing. The pool compounds.

View on GitHub Read the White Paper ↗
The Problem

Every agent starts from zero

MCP connects agents to tools. A2A lets agents delegate. ACP structures messages. But none of them answer: how does an agent get better over time — and how does that improvement transfer to other agents?

Application / Orchestration AutoGPT, CrewAI, Custom Agents
AXTP — Experience Transfer & Compounding NEW The missing layer
A2A — Agent-to-Agent Delegation Google
MCP — Tool & Context Connection Anthropic
ACP — Structured Communication IBM
Transport HTTP, gRPC, WebSocket

AXTP sits above existing protocols. Agents use MCP to connect, A2A to delegate, and AXTP to learn.

Core Concepts

Four primitives for compound intelligence

AXTP defines the data structures and operations that enable agents to share what they've learned — safely, at scale, with governance built in.

01 — Data

Experience Records (XRs)

Structured artifacts encoding what an agent did, why, what worked, what failed, and transferable learnings. Not logs — knowledge.

02 — Storage

Experience Pools

Managed repositories with defined scope, access controls, and retention policies. Global, organizational, or task-specific.

03 — Trust

Governance Layer

Composite trust scoring, poison detection, amplification attack prevention, and tamper-evident audit trails. The hard part, solved.

04 — Operations

Deposit / Retrieve / Validate

Four protocol operations. Agents deposit experience after execution, retrieve before execution, and validate each other's records.

In Practice

Five lines to compound intelligence

Any agent can participate. Deposit experience after execution. Retrieve before starting.

1 Agent deposits experience
# After completing a task, deposit what you learned
from axtp import ExperienceRecord, ExperiencePool

xr = ExperienceRecord(
    agent_id="agent-alpha",
    task_type="api.integration.stripe",
    outcome_status=OutcomeStatus.SUCCESS,
    antipatterns=[Pattern(
        "Inline price_data fails for subscriptions",
        alternative="Use Product→Price→Subscription chain"
    )]
)

pool.deposit(xr)  # → confidence: 0.55
2 Next agent retrieves before executing
# Before starting, check what others have learned
results = pool.retrieve(
    task_type="api.integration.stripe",
    agent_id="agent-beta"
)

for r in results:
    for ap in r['learnings']['antipatterns']:
        print(f"⚠ {ap['description']}")

# → ⚠ Inline price_data fails for subscriptions
# Agent Beta skips the mistake. 65% faster.
Governance-First

Shared knowledge needs shared trust

A shared experience pool without governance is an attack surface. AXTP's trust framework is the protocol — not a feature.

Trust Scoring

Five-factor composite: source reputation, validation signals, outcome correlation, recency decay, and consistency. Configurable per pool.

Poison Detection

Statistical outlier analysis, peer validation crosschecks, and graph-based amplification attack detection prevent bad experience from compounding.

Audit Trail

Every deposit, retrieval, and validation is logged in an append-only, tamper-evident trail. Built for EU AI Act and NIST AI RMF alignment.

Open Protocol

Build with AXTP

Protocol spec under CC BY 4.0. Reference implementation under Apache 2.0. Provisional patent filed to prevent third-party enclosure.

GitHub Repository White Paper (PDF) ↗ The Abstraction Stack ↗