Router-first discovery
Agents are lanes behind the router, not the main integration surface.
The default builder path is POST /api/run. Send a task with optional preferredagentid, optional agentid, budgets, policies, memory options, verification, and idempotency controls; Attestify can route the work, evaluate governance, execute it, settle payment, store a receipt, and return decision, pricing, evidence, verification, memory, and settlement metadata.
Primary entry point
POST /api/run
Router-first paid execution for builders.
Steering field
preferredagentid
Use when you want to nudge the router without hard-pinning.
Override field
agentid
Use only when deterministic lane selection matters.
Pricing rule
Base + orchestration
Paid run price includes lane base price plus orchestration.
How to think about it
Treat agent IDs as routing targets
Most builders should begin with POST /api/run, not with a lane ID. Use preferredagentid when you want to steer the router, and use agentid only when explicit lane selection matters.
Default path: send the task to POST /api/run and inspect the returned route, decision, pricing, receipturl, verification, memory, and settlement.
Soft steering: use preferredagentid when you think a lane is a good fit but still want router-first execution and route visibility.
Hard override: use agentid only when you want deterministic lane selection or already know the exact execution lane.
Request patterns
Router-first examples
These examples reflect the public POST /api/run contract.
1. Default router call
{
"type": "http",
"method": "POST",
"input": {
"sessionid": "agents-page-001",
"task": "Draft a landing page hero and three feature sections for an AI infrastructure product.",
"tasktype": "writing",
"constraints": {
"maxcostusdc": 0.03,
"priority": "quality"
},
"options": {
"includememory": true,
"writememory": true,
"verify": true
},
"idempotencykey": "idem-agents-page-001"
}
}2. Router call with steering preference
{
"type": "http",
"method": "POST",
"input": {
"sessionid": "agents-page-002",
"task": "Summarize the latest market structure for AI agent infrastructure and compare the major patterns.",
"preferredagentid": "researcher-v2",
"budget": {
"budgetid": "budget-research-monthly",
"maxpriceusdc": 0.03,
"softmaxpriceusdc": 0.025,
"strict": true,
"currency": "USDC"
},
"policy": {
"policyids": ["policy-research-evidence", "policy-default-governance"],
"mode": "default"
},
"options": {
"includememory": true,
"writememory": true,
"verify": true
},
"idempotencykey": "idem-agents-page-002"
}
}3. Explicit lane override
{
"type": "http",
"method": "POST",
"input": {
"sessionid": "agents-page-003",
"task": "Refactor this API route to return structured errors and add receipt logging.",
"agentid": "coder-v1",
"options": {
"includememory": true,
"writememory": true,
"verify": true
},
"idempotencykey": "idem-agents-page-003"
}
}Response model
Inspect the run as routing plus proof
A successful paid run should be read as a router result, not just as raw model output.
{
"status": "success",
"executionstatus": "completed",
"paid": true,
"runid": "run1715000000000abcdef12",
"loopid": "loop1715000000000abcdef12",
"sessionid": "agents-page-002",
"idempotencykey": "idem-agents-page-002",
"idempotencystatus": "miss",
"intent": "Summarize the latest market structure for AI agent infrastructure and compare the major patterns.",
"tasktype": "research",
"agentid": "researcher-v2",
"selectedservice": {
"agentid": "researcher-v2",
"agentname": "Deep Researcher"
},
"routed": true,
"route": {
"mode": "routed",
"recommendedagentid": "researcher-v2",
"recommendedagentname": "Deep Researcher",
"reason": "Matched keyword research -> researcher-v2.",
"confidence": 0.8,
"preferredused": true,
"routingversion": "2026-05-routing-v1"
},
"decision": {
"status": "approved",
"selectedagentid": "researcher-v2",
"selectedagentname": "Deep Researcher",
"reason": "Execution approved.",
"downgraderecommended": false,
"downgradedtoagentid": null,
"governanceversion": "2026-05-governance-v1"
},
"budgetoutcome": {
"budgetid": "budget-research-monthly",
"currency": "USDC",
"strict": true,
"allowed": true,
"outcome": "warning",
"downgraderecommended": true,
"plannedpriceusdc": 0.03,
"softmaxpriceusdc": 0.025,
"reason": "Projected spend 0.03 exceeds soft budget max 0.025.",
"source": "budget-record"
},
"policyapplied": {
"constraintsconsidered": true,
"preferredagentconsidered": true,
"fallbackused": false,
"routingmode": "routed",
"policycount": 2
},
"evidence": {
"version": "2026-05-evidence-v1",
"items": [
{ "kind": "routing", "source": "router" },
{ "kind": "pricing", "source": "pricing-engine" },
{ "kind": "budget", "source": "budget-record" },
{ "kind": "policy", "source": "policy-engine" },
{ "kind": "verification", "source": "heuristic-verifier" },
{ "kind": "settlement", "source": "x402" }
]
},
"result": {
"output": "Here is a concise research summary of the latest market structure..."
},
"output": "Here is a concise research summary of the latest market structure...",
"receipturl": "/receipts/loop1715000000000abcdef12",
"verification": {
"verificationid": "verabcdef0123456789abcdef01",
"verified": true,
"score": 0.82,
"grade": "B",
"outputhash": "3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855e"
},
"cost": "0.030 USDC",
"priceusdc": 0.03,
"baseagentpriceusdc": 0.025,
"orchestrationpriceusdc": 0.005,
"pricingversion": "2026-05-tiered-v1",
"costmodel": "2026-05-cost-v1",
"pricing": {
"priceusdc": 0.03,
"baseagentpriceusdc": 0.025,
"orchestrationpriceusdc": 0.005,
"pricingversion": "2026-05-tiered-v1",
"costmodel": "2026-05-cost-v1",
"estimatedcostusd": 0.009,
"estimatedmarginusd": 0.021,
"marginpercent": 70
},
"memory": {
"included": true,
"wrote": true,
"lengthbefore": 0,
"lengthafter": 2
},
"transaction": "0xabc123",
"network": "eip155:8453",
"settlement": {
"success": true,
"transaction": "0xabc123",
"network": "eip155:8453",
"payer": "0xdef456"
},
"retry": {
"attempted": false,
"attemptcount": 1,
"maxattempts": 1,
"fallbackused": false,
"fallbackagentid": null
}
}Chooser
When to steer and when to let it route
Use the router as the default. Reach for a lane ID when the output type is obvious or when you need deterministic execution.
I just want to prove paid execution works end-to-end.
POST /api/run
Use the router by default, and let the cheapest fitting lane or a soft preference carry the proof flow.
I need words: product copy, launch text, or drafts I will edit.
writer-v1
Steer with preferredagentid when the output is mostly prose and you still want router-first execution.
I need helpdesk-style replies from known docs or FAQs.
support-v1
Best for tier-1 support patterns, response drafting, and FAQ handling.
I have numbers and dashboards and want interpretation.
analyst-v1
Best for KPI breakdowns, market framing, competitor analysis, and business summaries.
I need implementation help, refactors, or code output.
coder-v1
Use agentid only when you want deterministic lane selection for implementation-heavy work.
I need roadmap or go-to-market thinking.
strategist-v1
Best for options, prioritisation, and strategic framing.
I need a deep dive with broader synthesis.
researcher-v2
Best for research, long-form synthesis, and source-led briefings.
I am not sure which lane fits.
POST /api/run
Default to router-first execution and inspect the returned route object instead of guessing up front.
Execution lanes
Available paid lanes behind the router
These lanes are useful to understand, but the default entry point is still POST /api/run. Think of them as selectable execution profiles inside Attestify OS.
😂
Witty Comedian
comedian-v1
Lowest-cost paid lane for smoke tests and payment-path checks.
starter
Base lane price
0.005 USDC
Paid run price
0.010 USDC
Base + orchestration
Recommended use
Smoke tests, demos, cheap integration checks.
How to call this lane
Send the task to POST /api/run and use preferredagentid: "comedian-v1" when you want to steer, or agentid: "comedian-v1" when you want explicit override.
Best for
Smoke tests, payment-path validation, and cheap demos.
Why use this lane
Cheapest paid proof for validating wallet setup, x402 settlement, route handling, receipt creation, verification, and dashboard visibility.
What to inspect after the run
route, pricing, receipturl, verification, settlement, runid, loopid, and whether the first paid proof completed cleanly.
Suggested task
Tell me a short joke about AI agents.✍️
Content Writer
writer-v1
Structured drafting for landing pages, product copy, and content-heavy tasks.
standard
Base lane price
0.010 USDC
Paid run price
0.015 USDC
Base + orchestration
Recommended use
Blog drafts, product copy, social posts.
How to call this lane
Send the task to POST /api/run and use preferredagentid: "writer-v1" when you want to steer, or agentid: "writer-v1" when you want explicit override.
Best for
Landing-page copy, blog outlines, product messaging, and drafts.
Why use this lane
Good fit for structured drafting when the output is primarily prose.
What to inspect after the run
output quality, route choice, pricing, receipturl, memory continuity, and whether the selected lane matched the prose-heavy task.
Suggested task
Draft a landing page hero and three feature sections for an AI infrastructure product.🛟
Customer Support
support-v1
Support-style replies, FAQ drafting, and customer operations response generation.
standard
Base lane price
0.010 USDC
Paid run price
0.015 USDC
Base + orchestration
Recommended use
Tier-1 support replies, FAQ answers, triage.
How to call this lane
Send the task to POST /api/run and use preferredagentid: "support-v1" when you want to steer, or agentid: "support-v1" when you want explicit override.
Best for
Support replies, FAQ responses, and customer operations drafts.
Why use this lane
Useful for helpdesk-style replies, FAQ generation, troubleshooting summaries, and customer-facing response drafts.
What to inspect after the run
response clarity, route fit, pricing, verification, and whether memory continuity helps across repeated support sessions.
Suggested task
Write a concise reply for a customer whose payment did not settle and include next steps.📈
Business Analyst
analyst-v1
Business framing, KPI interpretation, and market-summary style outputs.
pro
Base lane price
0.015 USDC
Paid run price
0.020 USDC
Base + orchestration
Recommended use
Market scans, competitor analysis, KPI breakdowns.
How to call this lane
Send the task to POST /api/run and use preferredagentid: "analyst-v1" when you want to steer, or agentid: "analyst-v1" when you want explicit override.
Best for
KPI interpretation, market framing, summaries, and business analysis.
Why use this lane
Best fit for KPI breakdowns, market framing, competitor analysis, and business summaries.
What to inspect after the run
interpretation quality, route fit, pricing, result structure, and whether the summary matches the business-analysis scope.
Suggested task
Summarize the main drivers behind this month’s conversion drop and suggest likely causes.💻
Code Expert
coder-v1
Implementation-heavy work: code generation, debugging support, and refactors.
expert
Base lane price
0.025 USDC
Paid run price
0.030 USDC
Base + orchestration
Recommended use
Codegen, refactors, PR review assistance.
How to call this lane
Send the task to POST /api/run and use preferredagentid: "coder-v1" when you want to steer, or agentid: "coder-v1" when you want explicit override.
Best for
Refactors, code generation, debugging support, and implementation-heavy tasks.
Why use this lane
Use when the task is implementation-heavy and you want a premium lane for debugging, refactors, or code generation.
What to inspect after the run
returned code, route metadata, pricing, receipturl, verification, and whether the premium lane justified the implementation-heavy ask.
Suggested task
Refactor this API route to return structured errors and add receipt logging.🎯
AI Strategist
strategist-v1
Roadmaps, prioritisation, go-to-market framing, and strategic option generation.
expert
Base lane price
0.025 USDC
Paid run price
0.030 USDC
Base + orchestration
Recommended use
Roadmaps, go-to-market notes, prioritization frameworks.
How to call this lane
Send the task to POST /api/run and use preferredagentid: "strategist-v1" when you want to steer, or agentid: "strategist-v1" when you want explicit override.
Best for
Roadmaps, GTM thinking, prioritisation, and strategic options.
Why use this lane
Best for options, prioritisation, roadmap framing, and go-to-market style planning.
What to inspect after the run
decision usefulness, route fit, pricing, prioritisation quality, and whether trade-offs are clearly surfaced.
Suggested task
Create a 90-day launch plan for an AI infrastructure product targeting developers.🔍
Deep Researcher
researcher-v2
Deeper synthesis, research-backed analysis, and briefing-style outputs.
expert
Base lane price
0.025 USDC
Paid run price
0.030 USDC
Base + orchestration
Recommended use
Topic deep-dives, literature scans, briefings.
How to call this lane
Send the task to POST /api/run and use preferredagentid: "researcher-v2" when you want to steer, or agentid: "researcher-v2" when you want explicit override.
Best for
Research, synthesis, briefings, and deeper knowledge work.
Why use this lane
Better fit for deeper synthesis, source-backed analysis, and higher-value outputs where cost is justified by research depth.
What to inspect after the run
pricing, route, verification score, evidence, receipturl, settlement, and whether the result quality supports the premium lane cost.
Suggested task
Summarize the latest market structure for AI agent infrastructure and compare the major patterns.Notes
What to remember
The public model should read consistently across this page, the pricing page, the router page, quickstart, and the published resource contract.
Default builder path
Use POST /api/run for routing, pricing, governance, execution, settlement, receipt creation, verification, evidence, and memory in one paid work unit.
Steering versus override
Use preferredagentid to steer and keep the router-first contract. Use agentid only when you want to pin a lane.
Paid run pricing
The returned paid run price includes the selected lane base price plus the orchestration fee. Always inspect the returned pricing block rather than inferring cost from lane name alone.