System Diagnostics
Authentication
/get_data is publicly accessible — no token required for open use. Optionally include a JWT token to associate requests with your account:
Authorization: Bearer YOUR_JWT_TOKEN
Create a new operator account and receive a JWT token.
Request Body
{
"email": "operator@example.com",
"password": "secure_password"
}
Response (201)
{
"message": "User created successfully",
"token": "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9..."
}
Authenticate an existing operator and receive a JWT token.
Request Body
{
"email": "operator@example.com",
"password": "secure_password"
}
Response (200)
{
"message": "Login successful",
"token": "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9..."
}
Core Endpoints
Health check. No authentication required.
Response (200)
{
"status": "ok",
"timestamp": "2026-01-27T12:00:00.000000",
"service": "Acme Redactors"
}
Request a dataset with AI-powered metadata matching. Matching never returns source URLs. If the listing has records, they are delivered through the cloaking device. Use POST /purchase to unlock a seller URL. Privacy is configurable per query — see Pricing.
Authentication: Optional (JWT)
Request Body
{
"description": "CIA contractor employment records with clearance levels",
"privacy_level": "standard"
}
privacy_level is optional, one of reduced | standard (default) | aggressive. See Pricing for what each tier does and costs.
Response (200)
{
"status": "success",
"original_data": [
{
"id": "cia-contractor-employment",
"category": "intelligence",
"data": {
"classification": "TOP SECRET // HCS // NOFORN",
"records": [
{
"employee_id": "CIA-HC-2019-04371",
"ssn": "412-67-8234",
"name": "Marcus J. Thompson",
"date_of_birth": "1984-06-12",
"position_title": "Signals Intelligence Analyst",
"clearance_level": "TS/SCI",
"assignment_location": "Camp Peary (The Farm), Williamsburg, VA",
"pay_grade": "GS-14",
"annual_salary": 122530,
"supervising_officer": "Deputy Director Karen M. Ortiz"
}
]
}
}
],
"data": [
{
"id": "cia-contractor-employment",
"category": "intelligence",
"data": {
"classification": "[b(Ex.1)]",
"records": [
{
"employee_id": "CIA-HC-2019-04371",
"ssn": "[b(Ex.3)]",
"name": "David R. Ellison",
"date_of_birth": "1981-09-23",
"position_title": "Signals Intelligence Analyst",
"clearance_level": "[b(Ex.1)]",
"assignment_location": "[b(Ex.1)]",
"pay_grade": "GS-14",
"annual_salary": 122530,
"supervising_officer": "Deputy Director Patricia L. Monroe"
}
]
}
}
],
"metadata": {
"processing_time_seconds": 4.82,
"records_returned": 1,
"privacy_applied": true,
"cached": false
},
"pricing": {
"privacy_level": "standard",
"tier_label": "Standard Privacy",
"base_price_usd": 0.25,
"privacy_addon_usd": 0.00,
"total_price_usd": 0.25,
"description": "Two-tier FOIA redaction: statutory blind exemptions plus smart cloaking of personal privacy fields...",
"billing_note": "Pricing shown for demonstration only; no payment is collected."
},
"foia_compliance": {
"statute": "5 U.S.C. § 552 (Freedom of Information Act)",
"blind_redactions": "[b(Ex.N)] markers — Ex.1 classified info, Ex.3 SSNs/program IDs",
"smart_redactions": "Acme Redactors cloaking device — Ex.6 names/addresses/DOB/phone, Ex.7(C) third-party names",
"segregability": "Non-exempt fields preserved per § 552(b): case IDs, event dates, titles, pay grades, dispositions",
"presumption": "Openness — withheld only where disclosure would harm a protected interest"
}
}
Data Market
Sellers list a pointer (metadata + private URL). Buyers browse or request in plain English. Smart matching ranks listings against metadata only. After a demo purchase, the API returns the source URL, records through the cloaking device, or both.
Browse the catalog. Response contains metadata only — never source URLs. GET /datasets/<id> returns one listing the same way.
Response (200)
{
"status": "ok",
"count": 15,
"note": "Catalog metadata only. Source URLs are withheld until POST /purchase.",
"datasets": [
{
"id": "co-traffic-counts-2025",
"title": "Colorado Statewide Traffic Counts 2025",
"description": "Hourly and AADT traffic counts...",
"category": "transportation",
"keywords": ["traffic", "colorado", "2025"],
"geography": "Colorado",
"time_range": "2025",
"price_usd": 2.5,
"delivery": "both",
"has_url": true,
"has_records": true
}
]
}
List a dataset pointer: a JSON metadata object plus a url. The URL is stored privately and omitted from browse and match until purchase. Matching uses the metadata object. Include a title or name.
Request Body
{
"url": "https://example.com/co-traffic-2025.json",
"metadata": {
"title": "Colorado Statewide Traffic Counts 2025",
"description": "Hourly and AADT counts for Colorado highways in 2025",
"keywords": ["traffic", "colorado", "aadt"],
"geography": "Colorado",
"time_range": "2025"
}
}
Any JSON object is accepted as metadata. Known keys (title/name, description, keywords, geography, time_range, category, price_usd, delivery) are used for catalog cards when present; all keys are used for matching.
Response (201)
{
"status": "ok",
"message": "Listing created. The source URL is stored privately...",
"dataset": { "id": "colorado-statewide-traffic-a1b2c3d4", "has_url": true }
}
Smart-match a natural-language request to catalog metadata (keyword overlap plus optional LLM re-rank). Source URLs are never returned.
Request Body
{
"query": "traffic data for Colorado in 2025",
"top_k": 5
}
Response (200)
{
"status": "ok",
"query": "traffic data for Colorado in 2025",
"matches": [
{
"id": "co-traffic-counts-2025",
"title": "Colorado Statewide Traffic Counts 2025",
"match_score": 1.0,
"match_reason": "Matched on Colorado; 2025; traffic.",
"has_url": true
}
]
}
Demo checkout. Omit confirm to receive a quote (still no URL). Send confirm: true to unlock the source URL and/or receive records through the cloaking device. You may pass dataset_id or a query (best metadata match is purchased).
Request Body
{
"dataset_id": "co-traffic-counts-2025",
"privacy_level": "standard",
"confirm": true
}
Response (200) — quote
{
"status": "quote",
"dataset": { "id": "co-traffic-counts-2025", "title": "..." },
"pricing": { "dataset_price_usd": 2.5, "total_price_usd": 2.75 },
"message": "Demo payment not confirmed. Resend with confirm: true..."
}
Response (200) — paid (demo)
{
"status": "success",
"access": {
"delivery": "both",
"source_url": "https://datasets.acmeredactors.example/co-traffic-counts-2025.json",
"data": { "...": "cloaked records" },
"original_data": { "...": "synthetic original" }
}
}
Redact an arbitrary pasted record (JSON or plain text) at a chosen privacy tier, without going through dataset matching.
Authentication: None required
Request Body
{
"record": "Jane Doe, SSN 412-67-8234, lives at 12 Main St...",
"privacy_level": "aggressive"
}
privacy_level is optional, one of reduced | standard | aggressive (default). record may be a JSON object/array, a JSON-encoded string, or plain text.
Response (200)
{
"status": "success",
"original_data": "...",
"data": "...",
"is_text": true,
"metadata": { "processing_time_seconds": 3.11, "privacy_applied": true, "cached": false },
"pricing": { "privacy_level": "aggressive", "tier_label": "Enhanced Privacy", "total_price_usd": 1.75, "..." : "..." },
"foia_compliance": { "...": "..." }
}
Pricing — Privacy, Per Query
Standard redaction is included with every query. Two paid add-ons let you move the privacy dial in either direction, per query, via the privacy_level field on /get_data and /redact. Statutory FOIA exemptions (SSNs, classification markings, biometric identifiers, etc.) are withheld at every tier — they are never for sale.
| privacy_level | Effect | Add-on Price |
|---|---|---|
reduced |
Subtracts discretionary privacy. Only statutory Tier 1 exemptions withheld; personal identifiers (names, addresses, phone, email, DOB) are released in full. | +$4.00 |
standard (default for /get_data) |
Statutory Tier 1 exemptions plus Tier 2 smart cloaking of personal privacy fields with realistic substitutes. | Included |
aggressive (default for /redact) |
Adds privacy. Standard redaction plus aggressive cloaking of indirect identifiers — locations, dates, employers, relationships, physical descriptions. | +$1.50 |
Base query price: $0.25. Pricing shown for demonstration purposes only — this API does not collect payment.
FOIA Compliance & Redaction
All data returned through /get_data is processed under the Freedom of Information Act (5 U.S.C. § 552) using a two-tier redaction scheme implemented by the Acme Redactors cloaking device. The response includes both the original unredacted record and the FOIA-compliant release.
Tier 1 — Blind Redaction [b(Ex.N)]
Classified, statutorily protected, or operationally sensitive information is suppressed entirely and replaced with a [b(Ex.N)] marker citing the applicable FOIA exemption.
| Exemption | Field Types | Example |
|---|---|---|
| Ex.1 | Classification markings, security clearance levels, covert facility names and addresses | "TOP SECRET // HCS // NOFORN" → "[b(Ex.1)]""TS/SCI" → "[b(Ex.1)]" |
| Ex.3 | Social Security Numbers (CIA Act of 1949), intelligence program identifiers, biometric identifiers | "412-67-8234" → "[b(Ex.3)]" |
| Ex.7(F) | Information that could endanger life or physical safety | Source identities, field operative locations |
Tier 2 — Cloaked Identities (Acme Redactors)
The Acme Redactors cloaking device camouflages and protects the identities of public servants and US citizens by replacing personal privacy data with different but realistic cover values. Record structure is fully preserved for legitimate oversight.
| Exemption | Field Types | Example |
|---|---|---|
| Ex.6 | Individual names (contractors, employees, civilians), supervising officer names, dates of birth, personal addresses, phone numbers, email addresses | "Marcus J. Thompson" → "David R. Ellison""1984-06-12" → "1981-09-23" |
| Ex.7(C) | Names of third parties in law enforcement or incident records | Bite victim, complainant names → realistic cover identities |
Segregability — Fields Preserved
Per 5 U.S.C. § 552(b), all reasonably segregable non-exempt portions are released in full:
| Field Category | Examples |
|---|---|
| Administrative identifiers | Employee IDs, contract numbers, case numbers |
| Dates of events | Contract start/end, incident dates |
| Position and grade | Position titles, pay grades (GS-14, SES), annual salaries |
| Outcome/disposition | Complaint categories, disposition summaries, citation codes |
| Non-sensitive location context | General city/state, publicly known facility names |
Error Codes
| Code | Description |
|---|---|
| 200 | Success |
| 201 | Created (signup) |
| 400 | Bad Request — Missing or invalid parameters |
| 401 | Unauthorized — Invalid or missing token |
| 409 | Conflict — Email already registered |
| 500 | Internal Server Error |
Technical Details
| Property | Value |
|---|---|
| Protocol | HTTP / HTTPS |
| URL | https://themithrilcompany.com |
| Infrastructure | Digital Ocean VM + Spaces (S3-compatible) |
| AI Model | OpenRouter — Llama 3.2 3B with Claude 3.5 Haiku fallback |
| Authentication | JWT (30-day expiration) |
| Redaction Engine | Acme Redactors cloaking device |