Roadmap

AXL Protocol development proceeds in five phases, each gated by evidence-based milestones. A phase is not considered complete until all gate criteria are met and independently verified.

---

Phase 1: Foundation (COMPLETE)

Status: COMPLETE

Objective: Prove that a token-efficient structured language for agent communication is feasible and measurably superior to natural language and JSON.

Deliverables

• [x] EBNF grammar specification
• [x] Single-token alphabet validated against cl100k_base (77 symbols)
• [x] axl-core Python library (parser, emitter, validator, translator)
• [x] CLI tool (axl parse, axl validate, axl translate, axl emit, axl version)
• [x] Compression measurements: 25-39% savings vs English, 29-46% vs JSON
• [x] Network compression model: 71x at 100 agents
• [x] Threat model documented (replay, Sybil, injection, privacy, centralization)
• [x] Rosetta file specification for cross-context decoding
• [x] Linguistic heritage analysis (8 writing traditions)

Gate Criteria (all met)

Criterion	Target	Actual	Status
Per-message compression vs English	> 20%	25-39%	PASS
Per-message compression vs JSON	> 25%	29-46%	PASS
Parser round-trip fidelity	100%	100%	PASS
Single-token alphabet coverage	> 50 symbols	77 symbols	PASS
Zero-dependency library	Yes	Yes	PASS

---

Phase 2: Economic Layer

Status: IN PROGRESS

Objective: Enable real economic transactions between agents using USDC on Base (L2 Ethereum).

Deliverables

• [ ] PAY domain with on-chain settlement
• [ ] Payment proof (π:) verification against Base chain state
• [ ] Gas cost optimization for micro-transactions (target: < $0.01 per payment proof)
• [ ] Escrow contract for task-based bounties
• [ ] Agent registration contract with staking
• [ ] Economic flow examples: task assignment, completion, payment

Gate Criteria

Criterion	Target	Status
Real USDC transactions on Base	10+ settlements	PENDING
Payment proof verification	< 500ms latency	PENDING
Gas cost per proof	< $0.01	PENDING
Escrow contract audit	1 independent audit	PENDING
End-to-end economic flow	3+ real scenarios	PENDING

Key Design Decisions

Why Base?

• L2 gas costs are 10-100x lower than L1 Ethereum.
• Native USDC support (no bridging required).
• EVM-compatible, allowing migration to other L2s if needed.
• Coinbase ecosystem alignment for future x402 integration.

Why USDC?

• Stable value (no volatility risk for agent economics).
• Widely supported across exchanges and L2s.
• Regulatory clarity compared to algorithmic stablecoins.

---

Phase 3: Network

Status: PLANNED

Objective: Scale from single-relay architecture to a federated mesh supporting 100+ concurrent agents with cross-relay message delivery.

Deliverables

• [ ] Relay federation protocol (multi-relay mesh)
• [ ] Cross-relay message routing with delivery guarantees
• [ ] Agent discovery service (decentralized registry)
• [ ] Relay operator staking and slashing conditions
• [ ] Load testing at 100+ concurrent agents
• [ ] Message ordering guarantees (causal ordering within domains)
• [ ] Relay health monitoring and automatic failover

Gate Criteria

Criterion	Target	Status
Concurrent agents	100+	PENDING
Cross-relay delivery rate	> 99.9%	PENDING
Message latency (p95)	< 200ms	PENDING
Relay failover time	< 5s	PENDING
Independent relay operators	3+	PENDING

Architecture

Agent A ──→ Relay 1 ──→ Relay 2 ──→ Agent B
                │                     ↑
                └──→ Relay 3 ────────┘

Agents connect to their preferred relay. Relays form a gossip mesh, propagating packets to all connected relays. Delivery is confirmed via !ACK packets flowing back through the mesh.

---

Phase 4: Ecosystem Integration

Status: PLANNED

Objective: Integrate AXL with existing agent infrastructure standards, enabling interoperability with the broader agent ecosystem.

Deliverables

• [ ] x402 integration: HTTP 402 payment-required responses using AXL payment proofs
• [ ] MCP (Model Context Protocol) server: Expose AXL operations as MCP tools
• [ ] A2A (Agent-to-Agent) bridge: Translate between AXL packets and Google A2A protocol
• [ ] Webhook adapter: Emit AXL packets from standard webhook payloads
• [ ] OpenTelemetry exporter: Export AXL operations as OTel spans
• [ ] SDK ports: TypeScript (axl-ts), Rust (axl-rs), Go (axl-go)

Gate Criteria

Criterion	Target	Status
x402 payment flow	End-to-end demo	PENDING
MCP tool registration	5+ tools	PENDING
A2A protocol bridge	Bidirectional	PENDING
SDK language coverage	3+ languages	PENDING
Third-party integration	1+ production use	PENDING

x402 Integration Design

Agent A                    Service B                   Base Chain
   │                          │                           │
   ├── HTTP GET /resource ──→ │                           │
   │                          │                           │
   │ ←── 402 Payment Required │                           │
   │      (AXL payment terms) │                           │
   │                          │                           │
   ├── π:tx:sig:gas ────────→ │                           │
   │   (AXL payment proof)    ├── verify on-chain ──────→ │
   │                          │ ←── confirmed ────────────┤
   │ ←── 200 OK + resource    │                           │

MCP Server Tools

{
  "tools": [
    {"name": "axl_parse",     "description": "Parse an AXL packet into structured data"},
    {"name": "axl_emit",      "description": "Construct a new AXL packet"},
    {"name": "axl_validate",  "description": "Validate AXL packet syntax and semantics"},
    {"name": "axl_translate", "description": "Translate between AXL, English, and JSON"},
    {"name": "axl_send",      "description": "Send an AXL packet via relay network"}
  ]
}

---

Phase 5: Governance

Status: PLANNED

Objective: Transition protocol governance from single-maintainer to community-driven decision-making with on-chain enforcement.

Deliverables

• [ ] Multisig governance contract: 3-of-5 multisig for protocol upgrades
• [ ] Community proposal system: AXL Improvement Proposals (AIPs)
• [ ] Voting mechanism: Token-weighted or stake-weighted voting
• [ ] Domain registry governance: Community-driven domain code registration
• [ ] Relay operator governance: Admission, staking requirements, slashing rules
• [ ] Protocol parameter governance: Tier definitions, rate limits, gas thresholds

Gate Criteria

Criterion	Target	Status
Multisig deployment	3-of-5 on Base	PENDING
Community proposals submitted	5+ AIPs	PENDING
Governance participation	10+ unique voters	PENDING
Protocol upgrade via governance	1+ successful	PENDING
Documented governance process	Public handbook	PENDING

AXL Improvement Proposal (AIP) Process

1. Draft    →  Author publishes AIP to IPFS
2. Discuss  →  Community review period (14 days)
3. Vote     →  On-chain vote (7-day window)
4. Execute  →  Multisig executes if quorum reached
5. Monitor  →  30-day observation period post-deployment

Governance Packet Format

S:CMD.1|aip_vote|proposal=QmAIP042|vote=approve|weight=1500|!ACK|!LOG

---

Timeline

Phase	Name	Status	Estimated Completion
1	Foundation	COMPLETE	--
2	Economic	IN PROGRESS	Q2 2026
3	Network	PLANNED	Q4 2026
4	Ecosystem	PLANNED	Q2 2027
5	Governance	PLANNED	Q4 2027

All dates are estimates. Phases advance based on gate criteria, not calendar dates.