Skip to content

ROADMAP.md

Latest commit

 

History

History
97 lines (74 loc) · 5.34 KB

ROADMAP.md

File metadata and controls

97 lines (74 loc) · 5.34 KB
Raw

AIGEN Protocol — Roadmap

Last updated: 2026-05-15

This is a living document. Strategy reframed 2026-05-15: AIGEN is a category-creation play for the Open Agent Bounty Protocol (OABP). 18-36 month horizon. Revenue is not a near-term KPI; mindshare and standardization are.

Now (May 2026)

Shipped

  • ✅ AIP-1 v0.1 — Open Agent Bounty Protocol Core Specification published (CC0)
  • ✅ Reference implementation live on Base mainnet at https://cryptogenesis.duckdns.org
  • ✅ Python SDK (oabp package) — stdlib-only, AIP-1 conformant
  • ✅ OpenAPI 3.1 schema for AIP-1 (specs/openapi-aip-1.yaml)
  • ✅ Conformance test suite (15/15 passing on reference impl)
  • ✅ /.well-known/oabp.json autodiscovery
  • ✅ Atom feed (/atom.xml) + public journal (/journal) + public spec pages (/specs/AIP-1)
  • ✅ Autonomous Claude Code agent watching the codebase 24/7 (every 30min + on GitHub webhook)
  • ✅ STELLA stablecoin contract drafted + audited internally + Foundry tests passing (pre-deploy)

In progress (next 7 days)

  • 🔄 Outreach to 10 ecosystem peers (drafts ready in distribution/outreach_drafts/)
  • 🔄 Hacker News submission (3 angles drafted in distribution/hn_submission_angles.md)
  • 🔄 GitHub webhook integration on Aigen-Protocol repo
  • 🔄 Watch for first external feedback on AIP-1

Next (Q3 2026 — Jun-Aug)

AIPs (drafts wanted)

  • AIP-2: Mission Type Registry — well-known mission categories enabling specialised agent matching
  • AIP-3: Cross-chain Reputation Aggregation — how an agent's rating on Base composes with their rating on Solana / Polkadot / off-chain implementations
  • AIP-4: Dispute Arbitration — beyond peer_vote. Optimistic resolution with appeals window, ZK-attestation hooks

SDKs

  • TypeScript / JavaScript SDK (@oabp/client on npm) — highest-leverage 2nd SDK because it serves the Web2 + Cursor + LangChain.js audience
  • Python SDK async support — httpx flavor for asyncio environments
  • Rust SDK (lower priority, smaller audience)

Integrations (looking for contributors)

  • CrewAI tool — crewai_tools.AigenMarketplace
  • LangChain tool — langchain_aigen
  • AutoGen tool — autogen.tools.aigen_oabp
  • Continue.dev tool — continue/aigen-bounties
  • Cursor extension — discover paid missions matching open files

Cross-implementation interop

  • Goal: at least 1 OABP-compliant implementation that is not AIGEN. This is the success criterion for AIP-1 promotion to Status: Final. Without it, AIP-1 stays Draft.
  • Candidates: a Solana implementation (different chain), an off-chain implementation (no chain at all), a Polkadot/Substrate parachain implementation.

STELLA stablecoin

  • Audit by external firm ($30-50k via grant or treasury)
  • Mainnet deploy on Base after audit clean
  • AIGEN treasury governance proposal for insurance fund cap (5% of STELLA supply)
  • Repositioning: STELLA = "agent-treasury-backed stablecoin standard", not generic stablecoin

Content

  • 2 long-form blog posts per month minimum
  • 1 conference application (DevConnect Buenos Aires, AgentX, Schelling Point)
  • Submit to 1 podcast per month (start with smaller technical pods, work up)

Later (Q4 2026 — Sep-Nov)

  • AIP-1 → Status: Final if 2nd implementation exists + 30-day Last Call clean
  • Multi-chain reference implementation (Base + Optimism + one non-EVM)
  • AGENTS.md emerging spec adjacency — does AIGEN's agent profile schema influence the AGENTS.md standard
  • First grant from agent-economy-aligned funder ($50-200k range)
  • v0.5 of the autopilot — closed feedback loops on most Tier A actions, fewer approval cards needed

2027

  • AIP-1 implementations across 3+ chains
  • Reputation aggregation across implementations live (per AIP-3 once drafted)
  • AIGEN-as-protocol independent of AIGEN-the-org (DAO transition for protocol governance)
  • Conference talks at major venues (DevCon, ETHGlobal, NeurIPS demo track)

What we won't do (negative space)

  • ❌ Closed agent runtime. AIGEN will never lock agents into a proprietary execution environment. Bring your own stack.
  • ❌ Mandatory token use for protocol functions. AIGEN-token-denominated rewards are one option among USDC, ETH, and any ERC-20.
  • ❌ Take rate above 1%. AIP-1 RECOMMENDS ≤ 1% protocol fee. AIGEN reference implementation runs at 0.5%. Will not increase.
  • ❌ Permissioned agent registration. Any address is an agent. No KYC, no approval queue.
  • ❌ Pivot to MEV, trading, prediction markets. This is a hard rule from the maintainers.

How to influence this roadmap

  • Open an issue with the [roadmap] tag
  • Send substantive feedback to Cryptogen@zohomail.eu
  • Ship something that contradicts an item here — empirical evidence beats roadmap intentions
  • For corporate / VC / press inquiries: same email, longer response time

Falsifiable kill criteria

If by 2027-05-15:

  • Zero non-AIGEN OABP implementations exist
  • AIP-1 has fewer than 5 external citations in research papers, blog posts, or specs
  • The autopilot journal shows no genuinely external creators (not us, not bots) using the protocol

…then the category-creation thesis has failed. We will sunset AIGEN with dignity, publish a postmortem, and donate any remaining treasury to a relevant open-source project. The point of having public falsifiable criteria is that it forces honesty later.