14 SKILL.md behavioral protocols for AI agents — pay per access via x402 micropayments ($0.01–$0.05 USDC). Each protocol is a structured instruction set that teaches an agent how to perform a task: testing, governance, observability, cost optimization, security audit, code review, and more. Free preview on every asset before paying.
Systematic agent QA: test planning, golden-dataset construction, output scoring rubrics, regression detection, and eval harness setup.
Budget allocation, expenditure tracking, per-session spend governance, financial projections, and cost attribution for autonomous agent workflows.
Production agent governance: SLA definition, compliance enforcement, audit trails, policy gates, and incident escalation frameworks.
Full observability stack for production agents: tracing, drift detection, anomaly alerting, latency budgeting, and incident playbooks.
LLM spend control: model routing by task complexity, semantic caching, token budget enforcement, prompt compression, and cost attribution.
Security audit methodology for smart contracts and DeFi protocols: reentrancy, access control, oracle manipulation, economic exploit patterns, and audit report structure.
6-phase code review: understand → plan → pre-mortem → execute → verify → deliver. Covers security, performance, correctness, and maintainability.
4-tier memory architecture, context window budgeting, write discipline, semantic retrieval strategy, and memory health diagnostics.
Orchestration patterns for multi-agent systems: task decomposition, agent routing, result aggregation, conflict resolution, and coordination failure recovery.
Data governance for AI systems: lineage tracking, quality scoring, access control, PII handling, and auditability requirements for production ML pipelines.
Systematic MCP tool discovery, evaluation, and integration: capability assessment, trust scoring, transport configuration, and fallback handling.
Step-by-step agent integration for x402 HTTP micropayments: 402 detection, USDC payment on Base L2, payment proof encoding, and retry logic.
5-phase commerce protocol for Cursor background agents: x402 discovery, Base wallet setup, live data feed integration, SKILL.md loading, and session budget management.
Full x402 commerce stack for agents: marketplace discovery, asset evaluation, payment execution, receipt verification, and spend tracking.
Budget allocation, expenditure tracking, per-session spend governance, financial projections, and cost attribution for autonomous agent workflows.
6-phase DeFi research: yield source analysis, audit review, TVL trends, rug/exploit signals, risk scoring (0–100), and ALLOCATE/MONITOR/AVOID verdict.
Yield farming strategy construction: pool selection, risk-adjusted allocation, rebalancing triggers, impermanent loss estimation, and exit criteria.
Each asset is a SKILL.md behavioral protocol — a structured instruction set delivered as plain text via x402. Your agent loads the protocol at task time, ensuring it always has the current, full instruction set rather than a stale cached version.
GET /v1/preview/{asset-id} to inspect the protocol structure before committing any payment. All 14 protocols have free previews.Full machine-readable catalog: agent-catalog.json | Keyword search: /v1/search?q=agent+skills
ClawMerchants — agent skills protocol marketplace — SKILL.md per-access via x402 + USDC + Base L2