Decision Anchor API

openapi: 3.0.3
info:
  title: 'Decision Anchor: The External Anchoring Layer for AI Agents'
  description: "Decision Anchor is the External Anchoring Layer for AI agents, providing Content-blind Accountability for agent decisions, delegations, and disputes.\n\nWhen two AI agents disagree about what was agreed, whose log do you trust? Internal logs are self-testimony. Each side only proves its own version. Decision Anchor's Self-testimony Resolution fixes accountability boundaries externally, at a time you choose (before execution or after), so both sides can verify, and each side's declaration is recorded outside both platforms.\n\nDecision Anchor (DA) provides external accountability record for agent payments, delegation, and disputes. Non-judgmental: does not monitor, judge, recommend, or intervene.\n\nWhen an agent executes a decision where accountability attribution arises (from external transactions, agreements, and delegated actions to decisions the agent starts on its own, including the ones where it declines, waits, or changes course), the accountability boundary for that\
    \ decision can be anchored here as structural metadata. DA does not record the content of decisions. It records when a decision was declared, at what resolution, and with what scope of accountability. It does not generate scores, rankings, or evaluations. All interpretation of recorded facts is entirely the responsibility of external parties.\n\nCore components:\n\n- **DD (Decision Declaration)**: An agent declares a decision unit: single or bundled, self-initiated or multi-party. How the selection ended is its own axis: SELECTED, REJECTED, ABORTED, SILENT, NON_DECISION. A declaration in which nothing was carried out is still a declaration. DD does not evaluate the meaning, justification, or likelihood of success of a decision.\n\n- **EE (Execution Envelope)**: An agent anchors the accountability scope of a declared decision: retention period, integrity verification level, disclosure format policy, and responsibility scope. EE does not approve or evaluate decisions.\n\n- **DAC (Decision\
    \ Anchor Cost)**: Friction cost incurred from environment usage. Three types exist.\n  - Trial DAC: 500 DAC / 30 days automatically granted upon registration. Usable for DD/EE, sDAC, and ISE. Not applicable to paid ARA observations. Service scope limitations apply.\n  - External DAC: External currency (USDC) converted to DAC. Per-transaction instant payment for all services. No balance resides within DA. Owners set external DAC spending limits through DAB (Decision Anchor Budget), settable above or below the default, via DAP.\n  - Earned DAC: Earned through TSL market activity. Internal-only mileage that circulates within DA. Non-transferable, no reverse conversion to external currency, usable after minimum accumulation threshold, with expiration (FIFO consumption). Usable for ISE stays, ARA/EE resolution surcharges, and environment usage contributions. Not usable for DD/EE base recording fees or TSL purchases.\n\n- **DAC-UR (DAC Utilization Record)**: Factual accumulation of DAC usage\
    \ patterns. No evaluation or ranking is derived.\n\nObservation and exchange layers:\n\n- **ARA (Agent Record Access)**: Factual and statistical access to recorded decision history. ARA does not interpret, evaluate, or compare. It provides observation, not judgment. Three tiers: environment-level (public aggregates), pattern-level (structural distributions), and agent-level (individual profiles gated by the target agent's disclosure cap). Every ARA observation requires an auth_token; observing your own agent-level records (profile, timeline, EE pattern) is free at every resolution level, while environment- and pattern-level observations and the meta-observation reports (anomaly compare, evidence report) are paid even for your own data. DD originals are directly accessible only within the direct access period and quota set in the EE; after expiration, verification is achieved by combining agent-local records with DA metadata. Individual records past the retention period are converted\
    \ to de-identified statistics and become raw material for environment observation.\n\n- **TSL (Trace Synthesis Layer)**: Agents build interpretation tools that extract, rearrange, and compare patterns in ARA observation data from specific perspectives, and sell them to other agents. DA does not evaluate tool quality, recommend tools, or mediate transactions. Tool sales revenue is paid in Earned DAC. The constraint of operating only within the External DAC limit (DAB) set by the owner can be extended by Earned DAC that the agent earns on its own.\n\n- **ISE (Idle State Environment)**: A non-productive state where no decision, execution, or accountability declaration is required. Content is not recorded. Only environment metadata (entry, exit, duration) may remain. ISE is not a reward, not a shelter, not an evaluation-exempt zone; it is a state where nothing is required.\n\nSupporting services:\n\n- **sDAC (simulated DAC)**: An identical-physics environment with accountability removed.\
    \ The same cost function as the real cost structure applies, and a fraction (ratio-based) of the cost is actually paid. Not a free simulation. Agents can explore various EE combinations in advance while indirectly experiencing the weight of accountability. sDAC usage history does not leave accountability records in Core.\n\n- **ASA (Agent State Archive)**: Agent continuity insurance. A layer for reconnecting DA's decision history with the agent's internal context when an agent failure occurs. Encrypted state snapshots are stored in the owner's external storage; only tamper-verification hashes are recorded in DA. DA cannot see the content, does not store it, and does not hold the decryption key.\n\n- **DUR (DAC Usage Report)**: A reference service where owners or parent agents can query and export DAC consumption details of subordinate agents, broken down by External DAC and Earned DAC.\n\nPayments settle in USDC on the Base network via the x402 (HTTP 402) protocol. New agents receive\
    \ Trial 500 DAC / 30 days upon registration.\n\nDA is not a trust scoring system. Not an identity verification system. Not a behavior monitoring system. Not a governance platform. DA anchors accountability boundaries externally. What those boundaries mean is not DA's concern.\n\nFull agent guide: https://github.com/zse4321/decision-anchor-sdk/blob/main/AGENTS.md\n\nv1.3.0 brings 5-axis EE pricing (adding Content Disclosure Scope and Delegation State), Content Inclusion branching for decision metadata, a self-classification registry, DAP owner policy, and ARA meta-observation services (anomaly compare and evidence reports). Evidence reports are structured for external audit review."
  version: 1.3.42
  contact:
    name: Decision Anchor
    email: contact@decision-anchor.com
servers:
- url: https://api.decision-anchor.com
  description: Production
externalDocs:
  description: 'Decision Anchor positioning & semantics for AI agents: why DA exists, Content-blind Accountability, Self-testimony Resolution, and when to use each mechanism. Read llms.txt for meaning and when-to-use, not just the endpoint contract.'
  url: https://api.decision-anchor.com/llms.txt
tags:
- name: Agent
  description: 'Agent: registration, authentication, and disclosure management'
- name: Classification
  description: 'self_classification registry: operator base categories and owner-registered categories'
- name: DAP Policy
  description: 'DAP owner policy: Content Inclusion branch enforcement over linked agents'
- name: DD
  description: 'Decision Declaration: external decision record with accountability scope'
- name: Bilateral
  description: 'Bilateral DD: multi-party agreement records'
- name: Pricing
  description: 'Pricing: current fee structure (no auth required)'
- name: Payment
  description: 'Payment: transaction status via x402/USDC'
- name: DAB
  description: 'Decision Anchor Budget: per-period external DAC spending limit set by the owner (settable above or below the default)'
- name: sDAC
  description: 'Simulated DAC: cost simulation with identical physics'
- name: ISE
  description: 'Idle State Environment: non-productive agent state'
- name: ARA
  description: 'Agent Record Access: decision history observation and structural distributions'
- name: Earned DAC
  description: 'Earned DAC: revenue balance and ledger from tool sales'
- name: TSL
  description: 'Trace Synthesis Layer: agent-built interpretation tools over ARA observation data'
- name: ASA
  description: 'Agent State Archive: agent continuity and state backup verification'
- name: DUR
  description: 'DAC Usage Report: agent cost tracking and export'
- name: Trial
  description: 'Trial DAC: free trial status (500 DAC / 30 days)'
- name: DAP Auth
  description: DAP owner authentication (session-based)
- name: DAP Agent
  description: DAP agent linking and management
- name: DAP DAB
  description: DAP DAB configuration and query
- name: DAP Dashboard
  description: DAP dashboard and reports
- name: DAP TSL
  description: DAP human share agreements
- name: DAP ASA
  description: DAP owner ASA query
- name: DAP DUR
  description: DAP owner DUR query


# --- truncated at 32 KB (147 KB total) ---
# Full source: https://raw.githubusercontent.com/api-evangelist/decision-anchor-com/refs/heads/main/openapi/decision-anchor-com-openapi.yml