Skip to content
Back to skills

2581 Datasets Debt 89d816b0

ASecurity

**Endpoint:** `/v2/accounting/od/debt_to_penny` **Frequency:** Daily **Date Range:** 1993-04-01 to present

  • 9 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added October 11, 2026
developmentpythonrustgoapisecurity

Works with

  • api

Security analysis

A100/100

Scanned October 11, 2026

npx -y skills add tools-only/X-Skills --skill 2581-datasets-debt_89d816b0 --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of 2581 Datasets Debt 89d816b0?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for 2581 Datasets Debt 89d816b0
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tools-only-2581-datasets-debt-89d816b0/badge)](https://www.skillsdirectory.com/skills/tools-only-2581-datasets-debt-89d816b0)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

SKILL.md
# Debt Datasets — U.S. Treasury Fiscal Data

## Debt to the Penny

**Endpoint:** `/v2/accounting/od/debt_to_penny`  
**Frequency:** Daily  
**Date Range:** 1993-04-01 to present

Tracks the exact total public debt outstanding each business day.

**Key fields:**
| Field | Type | Description |
|-------|------|-------------|
| `record_date` | DATE | Date of record |
| `debt_held_public_amt` | CURRENCY | Debt held by the public |
| `intragov_hold_amt` | CURRENCY | Intragovernmental holdings |
| `tot_pub_debt_out_amt` | CURRENCY | **Total public debt outstanding** |

```python
# Current national debt
resp = requests.get(
    "https://api.fiscaldata.treasury.gov/services/api/fiscal_service/v2/accounting/od/debt_to_penny",
    params={"sort": "-record_date", "page[size]": 1}
)
latest = resp.json()["data"][0]
print(f"As of {latest['record_date']}: ${float(latest['tot_pub_debt_out_amt']):,.2f}")

# Debt over the last year
resp = requests.get(
    "https://api.fiscaldata.treasury.gov/services/api/fiscal_service/v2/accounting/od/debt_to_penny",
    params={
        "fields": "record_date,tot_pub_debt_out_amt",
        "filter": "record_date:gte:2024-01-01",
        "sort": "-record_date"
    }
)
df = pd.DataFrame(resp.json()["data"])
df["tot_pub_debt_out_amt"] = df["tot_pub_debt_out_amt"].astype(float)
```

## Historical Debt Outstanding

**Endpoint:** `/v2/accounting/od/historical_debt_outstanding`  
**Frequency:** Annual  
**Date Range:** 1790 to present

Annual record of U.S. national debt going back to the founding of the republic.

**Key fields:**
| Field | Type | Description |
|-------|------|-------------|
| `record_date` | DATE | Year-end date |
| `debt_outstanding_amt` | CURRENCY | Total debt outstanding |

```python
# Full historical debt series
resp = requests.get(
    "https://api.fiscaldata.treasury.gov/services/api/fiscal_service/v2/accounting/od/historical_debt_outstanding",
    params={"sort": "-record_date", "page[size]": 10000}
)
df = pd.DataFrame(resp.json()["data"])
```

## Schedules of Federal Debt

**Endpoint:** `/v1/accounting/od/schedules_fed_debt`  
**Frequency:** Monthly  
**Date Range:** October 2005 to present

Monthly breakdown of federal debt by security type and component.

**Key fields:**
| Field | Type | Description |
|-------|------|-------------|
| `record_date` | DATE | End of month date |
| `security_type_desc` | STRING | Type of security |
| `security_class_desc` | STRING | Security class |
| `debt_outstanding_amt` | CURRENCY | Outstanding debt |

## Schedules of Federal Debt by Day

**Endpoint:** `/v1/accounting/od/schedules_fed_debt_daily`  
**Frequency:** Daily  
**Date Range:** September 2006 to present

Daily version of federal debt schedules with two data tables.

## Treasury Report on Receivables (TROR)

**Endpoint:** `/v2/debt/tror`  
**Frequency:** Quarterly  
**Date Range:** December 2016 to present

Federal agency compliance and receivables data. Also includes:
- `/v2/debt/tror/data_act_compliance` — 120 Day Delinquent Debt Referral Compliance Report

**Key fields:**
| Field | Type | Description |
|-------|------|-------------|
| `record_date` | DATE | Quarter end date |
| `funding_type_desc` | STRING | Type of funding |
| `total_receivables_delinquent_amt` | CURRENCY | Delinquent amount |

```python
# TROR data, sorted by funding type
resp = requests.get(
    "https://api.fiscaldata.treasury.gov/services/api/fiscal_service/v2/debt/tror",
    params={"sort": "funding_type_id"}
)
```

## Gift Contributions to Reduce the Public Debt

**Endpoint:** `/v2/accounting/od/gift_contributions`  
**Frequency:** Monthly  
**Date Range:** September 1996 to present

Records voluntary contributions from the public to reduce the national debt.

## Interest Expense on the Public Debt Outstanding

**Endpoint:** `/v2/accounting/od/interest_expense`  
**Frequency:** Monthly  
**Date Range:** May 2010 to present

Monthly interest expense broken down by security type.

**Key fields:**
| Field | Type | Description |
|-------|------|-------------|
| `record_date` | DATE | Month end date |
| `security_type_desc` | STRING | Security type |
| `expense_net_amt` | CURRENCY | Net interest expense |
| `expense_gross_amt` | CURRENCY | Gross interest expense |

```python
# Get total interest expense by month
resp = requests.get(
    "https://api.fiscaldata.treasury.gov/services/api/fiscal_service/v2/accounting/od/interest_expense",
    params={
        "fields": "record_date,expense_net_amt",
        "filter": "record_date:gte:2020-01-01",
        "sort": "-record_date"
    }
)
df = pd.DataFrame(resp.json()["data"])
df["expense_net_amt"] = df["expense_net_amt"].astype(float)
```

## Advances to State Unemployment Funds (Title XII)

**Endpoint:** `/v2/accounting/od/title_xii`  
**Frequency:** Daily  
**Date Range:** October 2016 to present

States and territories borrowing from the federal Unemployment Trust Fund.

**Key fields:**
| Field | Type | Description |
|-------|------|-------------|
| `record_date` | DATE | Date of record |
| `state_nm` | STRING | State name |
| `debt_outstanding_amt` | CURRENCY | Outstanding advance amount |

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.