API Reference

44 endpoints across 3 categories — every public HTTP endpoint exposed by the platform, with query parameters, response shape, and an example payload. Hand-curated; not auto-generated from route handlers.

Total

GET

POST

Read

Export

Admin

Read APIs

Read-only endpoints powering consoles, dashboards, and listings.

35 endpoints

GET/api/provenanceTry it →

Recent provenance events across the platform, newest first.

Query parameters

Name	Type	Required	Description	Example
limit	number	optional	Max events to return (1-200).	50
agent	string	optional	Filter to a single agent id.	30-knowledge-rag
study	string	optional	Filter to a single study id.	—

Response shape

{ events: Array<{ id: string; agent_id: string; action: string; study_id: string | null; created_at: string; provenance_hash: string }>; total: number }

Example response

{
  "events": [
    {
      "id": "pe_01HZX7ABCD",
      "agent_id": "30-knowledge-rag",
      "action": "draft_protocol_amendment",
      "study_id": "STU-204",
      "created_at": "2026-06-05T14:22:01.000Z",
      "provenance_hash": "sha256:9f1c..."
    }
  ],
  "total": 1
}

GET/api/staging/scorecardsAuthTry it →

Per-record review scorecards for the staging queue.

Query parameters

Name	Type	Required	Description	Example
study	string	optional	Filter to a study.	—
status	"pending"\|"signed"\|"rejected"	optional	Filter by lifecycle state.	pending

Response shape

{ scorecards: Array<{ record_id: string; agent_id: string; risk_score: number; citation_count: number; status: string }> }

Example response

{
  "scorecards": [
    {
      "record_id": "sr_01HZX9",
      "agent_id": "04-essential-documents",
      "risk_score": 0.18,
      "citation_count": 4,
      "status": "pending"
    }
  ]
}

GET/api/crons/statusTry it →

Latest run + next-run estimate for every scheduled cron job.

Query parameters

None.

Response shape

{ crons: Array<{ name: string; last_run_at: string | null; next_run_at: string | null; last_status: "ok"|"error"|"running" }> }

Example response

{
  "crons": [
    {
      "name": "rollup-cost-meter",
      "last_run_at": "2026-06-05T14:00:00.000Z",
      "next_run_at": "2026-06-05T15:00:00.000Z",
      "last_status": "ok"
    }
  ]
}

GET/api/cost-meter60 req/min/ipTry it →

Aggregate token + dollar spend across all agents for a window.

Query parameters

Name	Type	Required	Description	Example
window	"24h"\|"7d"\|"30d"	optional	Rolling window for the rollup.	24h

Response shape

{ window: string; tokens_in: number; tokens_out: number; usd_total: number; per_model: Record<string, { tokens: number; usd: number }> }

Example response

{
  "window": "24h",
  "tokens_in": 1284032,
  "tokens_out": 318411,
  "usd_total": 42.91,
  "per_model": {
    "claude-sonnet-4-7": { "tokens": 1041000, "usd": 30.12 }
  }
}

GET/api/cost-meter/trendTry it →

Hourly cost samples for a sparkline chart.

Query parameters

Name	Type	Required	Description	Example
hours	number	optional	Number of hourly buckets (1-168).	24

Response shape

{ points: Array<{ bucket_at: string; usd: number; tokens: number }> }

Example response

{
  "points": [
    { "bucket_at": "2026-06-05T13:00:00.000Z", "usd": 1.82, "tokens": 51200 },
    { "bucket_at": "2026-06-05T14:00:00.000Z", "usd": 2.04, "tokens": 58110 }
  ]
}

GET/api/agents/leaderboardTry it →

Top agents by activity, success rate, or cost.

Query parameters

Name	Type	Required	Description	Example
sort	"activity"\|"success"\|"cost"	optional	Sort key.	activity
limit	number	optional	Max rows (1-50).	10

Response shape

{ rows: Array<{ agent_id: string; name: string; runs: number; success_rate: number; usd_total: number }> }

Example response

{
  "rows": [
    { "agent_id": "30-knowledge-rag", "name": "Knowledge RAG", "runs": 412, "success_rate": 0.97, "usd_total": 12.41 }
  ]
}

GET/api/agents/[id]/activityTry it →

Recent activity timeline for a single agent.

Query parameters

Name	Type	Required	Description	Example
limit	number	optional	Max items (1-200).	20

Response shape

{ agent_id: string; events: Array<{ id: string; action: string; study_id: string | null; created_at: string }> }

Example response

{
  "agent_id": "30-knowledge-rag",
  "events": [
    { "id": "pe_01HZX7", "action": "draft_protocol_amendment", "study_id": "STU-204", "created_at": "2026-06-05T14:22:01.000Z" }
  ]
}

GET/api/agents/[id]/dossierTry it →

Credibility dossier: prompt versions, evals, model bindings.

Query parameters

None.

Response shape

{ agent_id: string; current_prompt_version: string; model_id: string; eval_score: number; eval_count: number; last_changed_at: string }

Example response

{
  "agent_id": "30-knowledge-rag",
  "current_prompt_version": "v17",
  "model_id": "claude-sonnet-4-7",
  "eval_score": 0.94,
  "eval_count": 32,
  "last_changed_at": "2026-05-28T09:11:00.000Z"
}

GET/api/agents/[id]/costsTry it →

Token and dollar spend rollup for a single agent.

Query parameters

Name	Type	Required	Description	Example
window	"24h"\|"7d"\|"30d"	optional	Rolling window.	7d

Response shape

{ agent_id: string; window: string; tokens: number; usd: number; runs: number }

Example response

{
  "agent_id": "30-knowledge-rag",
  "window": "7d",
  "tokens": 312041,
  "usd": 9.42,
  "runs": 87
}

GET/api/agents/compareTry it →

Side-by-side comparison of two or more agents.

Query parameters

Name	Type	Required	Description	Example
ids	string (comma-separated)	required	Agent ids to compare.	30-knowledge-rag,31-evaluation

Response shape

{ agents: Array<{ agent_id: string; name: string; eval_score: number; runs: number; usd_total: number }> }

Example response

{
  "agents": [
    { "agent_id": "30-knowledge-rag", "name": "Knowledge RAG", "eval_score": 0.94, "runs": 412, "usd_total": 12.41 },
    { "agent_id": "31-evaluation", "name": "Evaluation", "eval_score": 0.91, "runs": 305, "usd_total": 8.10 }
  ]
}

GET/api/studies/activityTry it →

Cross-study activity rollup, newest first.

Query parameters

Name	Type	Required	Description	Example
limit	number	optional	Max rows (1-100).	25

Response shape

{ studies: Array<{ study_id: string; protocol: string; last_event_at: string; pending_reviews: number }> }

Example response

{
  "studies": [
    { "study_id": "STU-204", "protocol": "ACME-PRO-204", "last_event_at": "2026-06-05T14:22:01.000Z", "pending_reviews": 3 }
  ]
}

GET/api/studies/[id]/snapshotTry it →

Snapshot view of one study: state, agents, pending reviews.

Query parameters

None.

Response shape

{ study_id: string; protocol: string; phase: string; status: string; agents_active: string[]; pending_reviews: number }

Example response

{
  "study_id": "STU-204",
  "protocol": "ACME-PRO-204",
  "phase": "II",
  "status": "active",
  "agents_active": ["30-knowledge-rag", "27-audit-trail"],
  "pending_reviews": 3
}

GET/api/sponsor-requests/recentTry it →

Most recent sponsor portal requests across all sponsors.

Query parameters

Name	Type	Required	Description	Example
limit	number	optional	Max rows (1-100).	20

Response shape

{ requests: Array<{ id: string; sponsor: string; subject: string; status: string; created_at: string }> }

Example response

{
  "requests": [
    { "id": "req_01HZX8", "sponsor": "Acme Therapeutics", "subject": "Protocol amendment v3", "status": "in_review", "created_at": "2026-06-05T11:02:00.000Z" }
  ]
}

GET/api/timelineTry it →

Cross-agent activity timeline for the operations console.

Query parameters

Name	Type	Required	Description	Example
from	string (ISO timestamp)	optional	Start of window.	2026-06-04T00:00:00Z
to	string (ISO timestamp)	optional	End of window.	—
limit	number	optional	Max events (1-500).	—

Response shape

{ events: Array<{ id: string; agent_id: string; action: string; study_id: string | null; created_at: string }> }

Example response

{
  "events": [
    { "id": "pe_01HZX7", "agent_id": "30-knowledge-rag", "action": "draft", "study_id": "STU-204", "created_at": "2026-06-05T14:22:01.000Z" }
  ]
}

GET/api/heatmapTry it →

Agent x hour activity heatmap data for the last N days.

Query parameters

Name	Type	Required	Description	Example
days	number	optional	Lookback window in days (1-30).	7

Response shape

{ cells: Array<{ agent_id: string; hour: number; runs: number }> }

Example response

{
  "cells": [
    { "agent_id": "30-knowledge-rag", "hour": 14, "runs": 12 },
    { "agent_id": "30-knowledge-rag", "hour": 15, "runs": 9 }
  ]
}

GET/api/risksTry it →

Risk register entries with mitigation status.

Query parameters

Name	Type	Required	Description	Example
severity	"low"\|"med"\|"high"	optional	Filter by severity.	high

Response shape

{ risks: Array<{ id: string; title: string; severity: string; mitigation: string; owner: string }> }

Example response

{
  "risks": [
    { "id": "RSK-12", "title": "PHI in agent prompt", "severity": "high", "mitigation": "Redaction filter pre-prompt", "owner": "security" }
  ]
}

GET/api/portalAuthTry it →

Sponsor portal home payload — open requests, recent signatures.

Query parameters

None.

Response shape

{ open_requests: number; recent_signatures: Array<{ id: string; subject: string; signed_at: string }> }

Example response

{
  "open_requests": 1,
  "recent_signatures": [
    { "id": "sig_01HZ", "subject": "Protocol amendment v2", "signed_at": "2026-06-04T16:00:00.000Z" }
  ]
}

GET/api/costsTry it →

Top-level cost analytics across agents and models.

Query parameters

Name	Type	Required	Description	Example
window	"24h"\|"7d"\|"30d"	optional	Rolling window.	30d

Response shape

{ window: string; usd_total: number; by_agent: Record<string, number>; by_model: Record<string, number> }

Example response

{
  "window": "30d",
  "usd_total": 1284.41,
  "by_agent": { "30-knowledge-rag": 412.10 },
  "by_model": { "claude-sonnet-4-7": 1020.30 }
}

GET/api/validationTry it →

CSV/CSA validation reports and qualification evidence summary.

Query parameters

None.

Response shape

{ reports: Array<{ id: string; title: string; status: "passed"|"failed"|"pending"; updated_at: string }> }

Example response

{
  "reports": [
    { "id": "VAL-204-IQ", "title": "Installation Qualification", "status": "passed", "updated_at": "2026-06-01T09:00:00.000Z" }
  ]
}

GET/api/searchTry it →

Unified search across studies, agents, and sponsor requests.

Query parameters

Name	Type	Required	Description	Example
q	string	required	Query string.	amendment
limit	number	optional	Max results per bucket (1-50).	—

Response shape

{ q: string; studies: Array<{ id: string; protocol: string }>; agents: Array<{ id: string; name: string }>; requests: Array<{ id: string; subject: string }> }

Example response

{
  "q": "amendment",
  "studies": [{ "id": "STU-204", "protocol": "ACME-PRO-204" }],
  "agents": [],
  "requests": [{ "id": "req_01HZX8", "subject": "Protocol amendment v3" }]
}

GET/api/complianceTry it →

Part 11 + ALCOA+ live compliance posture.

Query parameters

None.

Response shape

{ signatures_24h: number; staging_pending: number; alcoa: Record<string, "ok"|"warn"|"fail"> }

Example response

{
  "signatures_24h": 18,
  "staging_pending": 4,
  "alcoa": {
    "A": "ok", "L": "ok", "C": "ok", "O": "ok", "A2": "ok",
    "Complete": "ok", "Consistent": "ok", "Enduring": "ok", "Available": "ok"
  }
}

GET/api/digestTry it →

Daily morning brief — last 24h activity rollup.

Query parameters

Name	Type	Required	Description	Example
date	string (YYYY-MM-DD)	optional	Specific day to fetch.	2026-06-05

Response shape

{ date: string; sections: Array<{ title: string; bullets: string[] }> }

Example response

{
  "date": "2026-06-05",
  "sections": [
    { "title": "Signatures", "bullets": ["18 protocol amendments signed", "0 deviations"] }
  ]
}

GET/api/notificationsAuthTry it →

Unified notifications feed for the current user.

Query parameters

Name	Type	Required	Description	Example
unread	boolean	optional	Restrict to unread items.	true
limit	number	optional	Max items (1-100).	—

Response shape

{ items: Array<{ id: string; kind: string; title: string; created_at: string; read: boolean }> }

Example response

{
  "items": [
    { "id": "n_01HZ", "kind": "review_requested", "title": "Review STU-204 amendment", "created_at": "2026-06-05T14:00:00.000Z", "read": false }
  ]
}

POST/api/audit/queryAuthTry it →

Run a structured audit query against the provenance spine.

Query parameters

Name	Type	Required	Description	Example
format	"json"\|"sql"	optional	Return matching events or the SQL predicate.	json

Response shape

{ sql: string; rows: Array<{ id: string; agent_id: string; action: string; created_at: string }>; total: number }

Example response

{
  "sql": "SELECT * FROM cro.provenance_event WHERE agent_id = $1 AND created_at >= $2",
  "rows": [
    { "id": "pe_01HZX7", "agent_id": "30-knowledge-rag", "action": "draft", "created_at": "2026-06-05T14:22:01.000Z" }
  ],
  "total": 1
}

GET/api/audit/dailyTry it →

Per-day rollup of audit events for a window.

Query parameters

Name	Type	Required	Description	Example
days	number	optional	Lookback in days (1-90).	14

Response shape

{ days: Array<{ date: string; event_count: number; signature_count: number }> }

Example response

{
  "days": [
    { "date": "2026-06-04", "event_count": 412, "signature_count": 18 },
    { "date": "2026-06-05", "event_count": 380, "signature_count": 22 }
  ]
}

GET/api/audit/sankeyTry it →

Sankey-diagram-shaped flow data: agent then action then study.

Query parameters

Name	Type	Required	Description	Example
days	number	optional	Window length in days.	7

Response shape

{ nodes: Array<{ id: string; label: string }>; links: Array<{ source: string; target: string; value: number }> }

Example response

{
  "nodes": [
    { "id": "agent:30-knowledge-rag", "label": "Knowledge RAG" },
    { "id": "action:draft", "label": "draft" }
  ],
  "links": [
    { "source": "agent:30-knowledge-rag", "target": "action:draft", "value": 87 }
  ]
}

GET/api/ops-healthTry it →

Operational health: crons, endpoints, Supabase, SLO posture.

Query parameters

None.

Response shape

{ supabase: "ok"|"degraded"|"down"; crons_ok: number; crons_failing: number; slo_compliance: number }

Example response

{
  "supabase": "ok",
  "crons_ok": 12,
  "crons_failing": 0,
  "slo_compliance": 0.998
}

GET/api/status120 req/min/ipTry it →

Lightweight health check used by uptime monitors.

Query parameters

None.

Response shape

{ status: "ok"; version: string; ts: string }

Example response

{
  "status": "ok",
  "version": "1.42.0",
  "ts": "2026-06-05T14:30:00.000Z"
}

GET/api/flagsTry it →

Current feature-flag state catalog.

Query parameters

None.

Response shape

{ flags: Array<{ id: string; enabled: boolean; owner: string; last_toggled_at: string }> }

Example response

{
  "flags": [
    { "id": "agent-04-essential-documents", "enabled": true, "owner": "platform", "last_toggled_at": "2026-05-12T10:00:00.000Z" }
  ]
}

GET/api/dependenciesTry it →

Software bill of materials in JSON.

Query parameters

Name	Type	Required	Description	Example
category	string	optional	Filter by dependency category.	framework

Response shape

{ dependencies: Array<{ id: string; name: string; version: string; license: string; risk_class: string }> }

Example response

{
  "dependencies": [
    { "id": "nextjs", "name": "Next.js", "version": "15.0.0", "license": "MIT", "risk_class": "low" }
  ]
}

GET/api/modelsTry it →

Model registry: pinned model IDs and per-agent assignments.

Query parameters

None.

Response shape

{ models: Array<{ model_id: string; provider: string; pinned_at: string; bound_agents: string[] }> }

Example response

{
  "models": [
    { "model_id": "claude-sonnet-4-7", "provider": "anthropic", "pinned_at": "2026-05-01T00:00:00.000Z", "bound_agents": ["30-knowledge-rag", "27-audit-trail"] }
  ]
}

GET/api/meAuthTry it →

Current authenticated user identity and preferences.

Query parameters

None.

Response shape

{ user_id: string; email: string; role: string; mfa_enrolled: boolean; preferences: Record<string, unknown> }

Example response

{
  "user_id": "user_77",
  "email": "qa.lead@example.com",
  "role": "qa_lead",
  "mfa_enrolled": true,
  "preferences": { "theme": "light" }
}

GET/api/reports/[study]AuthTry it →

Generated report bundle for a single study.

Query parameters

Name	Type	Required	Description	Example
kind	"summary"\|"safety"\|"deviation"	optional	Which report flavor to return.	summary

Response shape

{ study_id: string; kind: string; generated_at: string; sections: Array<{ title: string; body: string }> }

Example response

{
  "study_id": "STU-204",
  "kind": "summary",
  "generated_at": "2026-06-05T14:30:00.000Z",
  "sections": [
    { "title": "Enrollment", "body": "42 subjects screened, 28 randomized." }
  ]
}

Export APIs

Bulk evidence downloads (CSV / JSON) for inspector handoff.

7 endpoints

GET/api/exports/agents.csvTry it →

CSV download of the full agent catalog.

Query parameters

None.

Response shape

text/csv stream — header row + one row per agent

Example response

agent_id,name,layer,model_id,eval_score
30-knowledge-rag,Knowledge RAG,L3,claude-sonnet-4-7,0.94

GET/api/exports/audit.jsonTry it →

Audit trail JSON bundle for a window.

Query parameters

Name	Type	Required	Description	Example
from	string (ISO timestamp)	optional	Start of export window.	—
to	string (ISO timestamp)	optional	End of export window.	—

Response shape

{ exported_at: string; events: Array<{ id: string; agent_id: string; action: string; created_at: string; provenance_hash: string }> }

Example response

{
  "exported_at": "2026-06-05T14:30:00.000Z",
  "events": [
    { "id": "pe_01HZX7", "agent_id": "30-knowledge-rag", "action": "draft", "created_at": "2026-06-05T14:22:01.000Z", "provenance_hash": "sha256:9f1c..." }
  ]
}

GET/api/exports/cron-history.csvTry it →

CSV history of every scheduled cron run.

Query parameters

Name	Type	Required	Description	Example
days	number	optional	Lookback window in days.	30

Response shape

text/csv stream — header row + one row per run

Example response

name,started_at,finished_at,status
rollup-cost-meter,2026-06-05T14:00:00Z,2026-06-05T14:00:03Z,ok

GET/api/exports/models.csvTry it →

CSV of the model registry — pinned model ids and assignments.

Query parameters

None.

Response shape

text/csv stream — header row + one row per model registry version

Example response

model_id,provider,bound_agents,pinned_at
claude-sonnet-4-7,anthropic,30-knowledge-rag,2026-05-01

GET/api/exports/provenance.csvTry it →

CSV of the provenance event spine for a window.

Query parameters

Name	Type	Required	Description	Example
from	string (ISO timestamp)	optional	Start of window.	—
to	string (ISO timestamp)	optional	End of window.	—

Response shape

text/csv stream — header row + one row per event

Example response

id,agent_id,action,study_id,created_at,provenance_hash
pe_01HZX7,30-knowledge-rag,draft,STU-204,2026-06-05T14:22:01Z,sha256:9f1c

GET/api/exports/signatures.csvAuthTry it →

CSV of every Part 11 e-signature row in the window.

Query parameters

Name	Type	Required	Description	Example
from	string (ISO timestamp)	optional	Start of window.	—
to	string (ISO timestamp)	optional	End of window.	—

Response shape

text/csv stream — header row + one row per signature

Example response

id,user_id,record_id,intent,signed_at
sig_01HZ,user_77,sr_01HZX9,Approve protocol amendment,2026-06-05T14:25:00Z

Admin APIs

Operator-only endpoints — gated behind admin role + MFA.

2 endpoints

POST/api/admin/seedAuth1 req/min/userTry it →

Seed the database with demo fixtures (admin only).

Query parameters

Name	Type	Required	Description	Example
reset	boolean	optional	Truncate demo rows first.	false

Response shape

{ inserted: number; skipped: number; duration_ms: number }

Example response

{
  "inserted": 412,
  "skipped": 0,
  "duration_ms": 1842
}

GET/api/admin/envAuthTry it →

Redacted environment variable catalog (admin only).

Query parameters

None.

Response shape

{ vars: Array<{ name: string; present: boolean; preview: string | null }> }

Example response

{
  "vars": [
    { "name": "ANTHROPIC_API_KEY", "present": true, "preview": "sk-...cd12" },
    { "name": "SUPABASE_URL", "present": true, "preview": "https://clkifpedvomnpgajdsid..." }
  ]
}