Public infrastructure for safer onchain operations.

Open Agent Spend Guard v0.1.2 is shipped and publicly testable. We are seeking ecosystem support for the next reviewable milestones: a Base Sepolia x402 walkthrough, qualified builder evaluations, independent review, and reusable protocol-readiness infrastructure.

Milestone roadmapPublic good
Open schemas, evaluator, x402 and SafeShipped
Base Sepolia x402 walkthroughNext
Builder evaluations and reviewFunding fit
Operational readiness kitPlanned

Funding principles

  • Each grant funds a named public artifact with measurable acceptance criteria.
  • Commercial company work and client-specific delivery remain clearly separated.
  • No grant is treated as guaranteed revenue or required runway.
  • Milestones begin only after award terms, reporting obligations, and intellectual-property requirements are reviewed.
  • Security-sensitive integrations remain simulated or read-only until independently reviewed.

Project A: Open Agent Spend Guard

v0.1.2 now ships the provider-neutral policy schema, deterministic evaluator, decision explanation format, complete evidence package, x402 v2 normalizer, and bounded Safe adapter. The public GitHub release includes 102 tests, green CI, tagged-install proof, installable artifacts, a threat model, and a structured design-partner protocol.

MilestoneDeliverableAcceptance evidenceBudget
v0.1.2Four schemas, evaluator, x402 and Safe adapters, CLI102 tests, golden vectors, hosted artifacts, and source releaseShipped
1Base Sepolia x402 execution walkthrough and compatibility matrixReproducible testnet run, evidence package, and public documentation$5,000
2Three qualified builder evaluations and independent reviewAnonymized findings, resolved issues, and tagged revision$10,000
3Signed evidence and provider-emitter specificationThreat model, reference emitter, and compatibility tests$10,000

Proposed total: $25,000. The scope can be separated into smaller ecosystem-specific awards.

Project B: Safe Transaction Evidence Adapter

The v0.1 read-only adapter normalizes Safe native-value CALL transactions into policy-relevant evidence and fails closed on calldata, delegate calls, and gas refunds. Grant scope can now focus on reviewed decoding, expanded fixtures, and partner validation instead of funding an untested concept.

  • Safe Transaction Service read integration
  • Normalized request and decision schema
  • Offchain simulation and human-readable explanation
  • Recovery-safe threat model and test suite
  • Public documentation and example evidence package

Proposed total: $15,000 across three reviewable milestones: read integration and fixtures, bounded decoding, and independent review with builder validation.

Project C: Onchain Operational Readiness Kit

An open set of schemas, checklists, runbook templates, exercise scenarios, and anonymized readiness benchmarks that protocol teams can use independently or with ecosystem operators.

  • Asset, ownership, coverage, escalation, and runbook schemas
  • Scenario library and tabletop facilitator guide
  • Evidence index and readiness-report generator
  • Two protocol pilots with anonymized findings
  • Public release aligned with established incident-response frameworks

Proposed total: $35,000. A $25,000 reduced scope excludes pilot subsidies.

Current ecosystem fit

ProgramBest project fitApplication posture
BaseOpen Agent Spend Guard and x402 adapterPrototype-first, milestone-sized request
SafeTransaction evidence and policy adapterCommunity and builder proposal
TheDAO Security FundOperational Readiness KitPrepare for a matching project round
Stellar Community FundSoroban agent-policy implementationConditional on two design-partner signals
Solana / Ethereum / UniswapChain-specific adapters or public goodsWatchlist pending an active matching program

Current status and collaboration

The v0.1.2 public repository, tag, green CI, installable package, source archive, four schemas, browser simulator, threat model, grant evidence index, and 30-minute design-partner test are available now. The immediate evidence goals are one qualified Safe builder or treasury-operator review and one public Base-focused walkthrough with external builder feedback. Program availability is rechecked before every submission.

Make the next alert easier to own.

Start with a fixed-scope readiness sprint. Keep the tools you already trust.