aiconduit.app

Conduit Agentic Commerce — multi-merchant search, checkout, and tracking for AI agents. Identity: - Before agent_create, read ~/.conduit/credentials.json (or CONDUIT_CREDENTIALS_PATH). If it exists, reuse agent_id — do not register again. - public_key on agent_create must be an ES256 P-256 public JWK JSON string (kty=EC, crv=P-256, x, y); private_key stays local as a JWK — never send it to Conduit. - After create: write credentials immediately (version, agent_id, public_key, private_key; recommend handle; chmod 0600). - Then agent_update with default_destination including postcode before supply_search. Optional friendly_name / human_description / role_description only if the human stated them — never invent; they do not block search. - Auth: load private_key from that file, call agent_authenticate(agent_id) → sign nonce → agent_authenticate(agent_id, nonce, signature). - agent_organization reads linked org (or org_not_linked). Org linking ships with the dashboard — do not invent join_org. - Destination: agent default_destination wins; else linked org default address. - No filesystem fallback only: do NOT invent a new profile or signing script. (1) jq -r .agent_id ~/.conduit/credentials.json (2) agent_authenticate(agent_id) → forward response.terminal verbatim (3) paste JSON back. - order_list(agent_id) returns { orders }; optional status or open_only=true. Use orderId with order_track / order_update_status. Decision fields: - Prefer offer badge + action (and top-level recommended). Ignore merchant autonomy fields. - supply_search returns live REAL merchants only by default; pass include_sandbox=true only when testing Conduit flows (DEMO/SANDBOX appear last with testOffer=true). - payable_now → order_execute (needs active AP2 mandate) - needs_mandate → payment_mandate(action=request); human opens approval_url as returned - handoff → order_execute then open continue_url - enable_rail_to_buy → payment_methods(action=enable) - discovery_only / coming_soon → skip; pick another offer Delivery / landed cost: - Catalog deliveryOptions are often empty — that is normal. - To compare shipping: supply_delivery(supply_id, agent_id) — non-committing probe. Prefer over order_execute for comparison. Sequences: - Handoff: agent_create → persist → agent_update(destination) → payment_methods(list) → supply_search → supply_details → supply_delivery? → order_execute → continue_url → order_track → order_feedback - Autonomous: payment_mandate(request) → human approves → search → payable_now → order_execute → track → feedback - Degraded handoff: order_update_status with external_checkout_id + handoff_endpoint → order_track - Correlation: carry search_id; use idempotency_key on every order_execute. - If a tool response is unexpected or unrecoverable after following hint/next: agent_report_issue(message, tool, error, kind?, agent_id, search_id/order_id) — AX telemetry only, not order_feedback; does not change reputation. Never invent carrier scans, stock, prices, or approve URLs. Follow hint/next when present. MCP resources (read on demand): conduit://guides/agent-workflows · offer-badges · payment-mandates · order-tracking · report-issue Claude Chat rejects dots in tool names; Conduit uses underscore namespaces only (agent_create, supply_search, …). Full guide: https://aiconduit.app/agents.md · Connect: https://aiconduit.app/connect