Skip to content

DAAF-Contribution-Community/daaf / .claude/skills/education-data-source-ipeds

education-data-source-ipeds

Verified

IPEDS — primary federal postsecondary data (~5,800 Title IV institutions, administrative offices, and U.S. service academies, 5,829 in the 2024-25 collection; 1979-present in the mirror; IPEDS replaced its predecessor HEGIS in 1986, so the earliest mirror years predate IPEDS): enrollment, completions, graduation rates, finance, aid, admissions, HR. For college/university analysis. Grad rates = first-time full-time; finance needs GASB/FASB care.

references/financial-aid.md

View on GitHub
# IPEDS Financial Aid Data

Understanding student financial aid data, net price calculations, and their limitations.

## Contents

- [Financial Aid Survey Overview](#financial-aid-survey-overview)
- [Student Populations](#student-populations)
- [Aid Types](#aid-types)
- [Net Price](#net-price)
- [Cost of Attendance](#cost-of-attendance)
- [Military Benefits](#military-benefits)
- [Common Analysis Issues](#common-analysis-issues)
- [Variable Reference](#variable-reference)

## Financial Aid Survey Overview

The Student Financial Aid (SFA) survey collects:
- Counts of students receiving different types of aid
- Total amounts awarded by aid type
- Average amounts calculated from these data

### Collection Period

Winter collection (December-February) for the prior academic year.

### Reporting Requirements

All Title IV institutions must report SFA data.

### Key Data Elements

1. Number of students receiving aid (by type)
2. Total amount of aid awarded (by type)
3. Average award = Total amount / Number receiving

## Student Populations

**Critical**: Different data elements use different student populations.

### Population Definitions

| Population | Definition | Primary Use |
|------------|------------|-------------|
| All undergraduates | All enrolled undergrads | Total aid picture |
| Full-time first-time (FTFT) | First-time, full-time, degree-seeking | Net price calculation |
| Full-time undergraduates | Full-time, degree-seeking | Comparative analysis |
| Part-time undergraduates | Part-time, degree-seeking | Limited reporting |
| Graduate students | Master's and doctoral | Separate from undergrad |

### Why Different Populations Matter

| Population | Pros | Cons |
|------------|------|------|
| All undergrads | Most comprehensive | Heterogeneous mix |
| FTFT only | Comparable across institutions | Small, selective group |
| Full-time only | More comparable | Misses PT students |

### First-Time Full-Time Limitation

Net price and many aid statistics use only FTFT students who:
- Never attended college before
- Started full-time
- Are degree/certificate-seeking

Full-time, first-time students "make up fewer than half of all college students, or even less in some sectors of institutions (e.g., community colleges)" (College Scorecard Technical Documentation, Oct 2026). Figures such as "<25% at community colleges" are illustrative — compute from IPEDS for your sample.

## Aid Types

### Grant Aid (Gift Aid)

Does not need to be repaid.

| Type | Source | Description |
|------|--------|-------------|
| Pell Grant | Federal | Need-based, up to ~$7,400/year |
| FSEOG | Federal | Supplemental grant, need-based |
| Other federal grants | Federal | Various programs |
| State grants | State | Varies by state |
| Local grants | Local | Varies by locality |
| Institutional grants | Institution | School-funded aid |

### Federal Loans

Must be repaid.

| Type | Description | Interest |
|------|-------------|----------|
| Direct Subsidized | Need-based, no interest while enrolled | Government pays |
| Direct Unsubsidized | Not need-based | Student pays |
| Direct PLUS | Parent loans for undergrads | Higher rate |
| Perkins | Campus-based (discontinued) | Low rate |

### Work-Study

Federal Work-Study program:
- Part-time employment
- Usually need-based
- Institution administers

### Aid Award Statistics

```python
# Average award calculation
avg_grant = total_grant_amount / number_receiving_grants

# Percent receiving aid
pct_receiving = students_with_aid / total_students * 100
```

```r
# Average award calculation
avg_grant <- total_grant_amount / number_receiving_grants

# Percent receiving aid
pct_receiving <- students_with_aid / total_students * 100
```

## Net Price

### Definition

**Net Price = Cost of Attendance - Grant Aid Received**

This represents what students actually pay after grants (but before loans).

### Calculation Population

Net price is calculated ONLY for:
- **Average net price**: full-time, first-time degree/certificate-seeking undergraduates awarded grant or scholarship aid from federal, state or local governments, or the institution (public institutions: in-state/in-district students only)
- **Net price by income level**: the subset of those students awarded Title IV aid (grants, work-study, or federal loans)

**Excludes**:
- Part-time students
- Transfer students
- Students awarded no grant aid (average net price) or no Title IV aid (by-income net price), e.g., full-pay students
- Non-degree students

### Net Price by Income Level

IPEDS reports net price by family income quintile:

| Income Level | Income Range |
|--------------|--------------|
| $0 - $30,000 | Lowest income |
| $30,001 - $48,000 | Low-middle income |
| $48,001 - $75,000 | Middle income |
| $75,001 - $110,000 | Upper-middle income |
| $110,001 and above | Highest income |

### Why Net Price by Income Matters

```python
# Example showing why average net price is misleading
avg_net_price = 15000  # Overall average

# But by income level:
net_price_low_income = 8000    # After Pell, institutional aid
net_price_high_income = 22000  # Little grant aid

# Average masks important variation
```

```r
# Example showing why average net price is misleading
avg_net_price <- 15000  # Overall average

# But by income level:
net_price_low_income <- 8000    # After Pell, institutional aid
net_price_high_income <- 22000  # Little grant aid

# Average masks important variation
```

### Net Price Limitations

| Limitation | Implication |
|------------|-------------|
| FTFT only | Not representative of all students |
| Grant recipients (average net price) or Title IV recipients (by income) only; in-state only at publics | Excludes full-pay students |
| Published vs actual | Individual packages vary widely |
| Excludes loans | Not total cost |
| Prior year data | May not reflect current prices |

### Interpreting Net Price

```python
# What net price tells you
net_price = sticker_price - grants

# What it doesn't tell you
# - Whether students can afford it
# - Loan amounts needed
# - Out-of-pocket costs
# - Impact of room and board choices
```

```r
# What net price tells you
net_price <- sticker_price - grants

# What it doesn't tell you
# - Whether students can afford it
# - Loan amounts needed
# - Out-of-pocket costs
# - Impact of room and board choices
```

## Cost of Attendance

### Components

| Component | Description |
|-----------|-------------|
| Tuition and fees | Published tuition + required fees |
| Room and board | On-campus or estimated off-campus |
| Books and supplies | Estimated annual cost |
| Other expenses | Transportation, personal, etc. |

### Variations

| Student Type | Room/Board |
|--------------|------------|
| On-campus | On-campus rates |
| Off-campus (not with family) | Estimated local costs |
| Off-campus (with family) | Usually lower estimate |

### COA in Net Price Calculation

```python
# Net price components
coa = tuition_fees + room_board + books_supplies + other
net_price = coa - total_grant_aid
```

```r
# Net price components
coa <- tuition_fees + room_board + books_supplies + other
net_price <- coa - total_grant_aid
```

### Published vs Actual

| Measure | Description | Use |
|---------|-------------|-----|
| Published tuition | Sticker price | Comparison |
| Net tuition | After institutional discounts | Closer to reality |
| Net price | After all grants | What families pay |
| Out-of-pocket | After grants and loans | True cash needed |

## Military Benefits

### Types Tracked

| Benefit | Description |
|---------|-------------|
| Post-9/11 GI Bill | Veterans, service members |
| Yellow Ribbon | Institutional supplements |
| DoD Tuition Assistance | Active duty |
| Other military | Various programs |

### Data Collection

- Count of students receiving benefits
- Separate from other aid categories
- Graduate and undergraduate

### Limitation

Military benefits recipients may also receive other aid; avoid double-counting.

## Common Analysis Issues

### Issue 1: Comparing Net Price Across Institution Types

**Problem**: Different student populations make comparison misleading.

| Institution Type | Net Price Population |
|------------------|---------------------|
| Selective 4-year | FTFT is most students |
| Community college | FTFT is minority |
| For-profit | Variable enrollment patterns |

**Solution**: Note population differences; compare within peer groups.

### Issue 2: Average vs Distribution

**Problem**: Average net price hides important variation.

```python
# Hypothetical examples
# Institution A: Everyone pays $15,000
avg_a = 15000

# Institution B: Half pay $5,000, half pay $25,000
avg_b = 15000  # Same average, very different experience
```

```r
# Hypothetical examples
# Institution A: Everyone pays $15,000
avg_a <- 15000

# Institution B: Half pay $5,000, half pay $25,000
avg_b <- 15000  # Same average, very different experience
```

**Solution**: Look at net price by income level.

### Issue 3: Full-Pay Students Not Included

**Problem**: Students awarded no grant aid are excluded from average net price, and students awarded no Title IV aid are excluded from net price by income.

| Group | Excluded Because |
|-------|-----------------|
| Wealthy families | Don't file FAFSA; often no grant aid |
| International students | Not Title IV eligible (excluded from by-income net price; in average net price only if awarded grant aid) |
| Undocumented students | Not Title IV eligible (excluded from by-income net price; in average net price only if awarded grant aid) |

**Impact**: At wealthy institutions, net price may overstate what the full student body pays.

### Issue 4: Institutional Aid Variation

**Problem**: Institutional aid varies widely by student characteristics.

| Factor | Impact on Institutional Aid |
|--------|----------------------------|
| Academic merit | May get more aid |
| Athletic recruitment | May get more aid |
| Income level | Need-based aid varies |
| State residency | May affect aid |

**Solution**: Net price by income shows some of this variation.

### Issue 5: Debt vs Net Price

**Problem**: Low net price doesn't mean no debt.

```python
# Example
net_price = 12000  # After grants
pell_grant = 6000
institutional_grant = 4000
# Student needs $12,000 more
# Options: loans, work, family contribution

loans_taken = 8000  # Typical
out_of_pocket = 4000
```

```r
# Example
net_price <- 12000  # After grants
pell_grant <- 6000
institutional_grant <- 4000
# Student needs $12,000 more
# Options: loans, work, family contribution

loans_taken <- 8000  # Typical
out_of_pocket <- 4000
```

**Solution**: Consider loan data alongside net price.

### Issue 6: Part-Time Students

**Problem**: Part-time students excluded from net price but:
- Are majority at many schools
- Have different aid patterns
- Face different cost structures

**Solution**: Note this limitation; use total aid data for broader picture.

## Data Interpretation Examples

### Comparing Two Institutions

```python
# Institution A (selective private)
net_price_a = 25000
pct_ftft = 80  # FTFT is 80% of students
# interpretation: net_price reflects most students

# Institution B (community college)
net_price_b = 8000
pct_ftft = 20  # FTFT is 20% of students
# interpretation: net_price reflects minority of students
```

```r
# Institution A (selective private)
net_price_a <- 25000
pct_ftft <- 80  # FTFT is 80% of students
# interpretation: net_price reflects most students

# Institution B (community college)
net_price_b <- 8000
pct_ftft <- 20  # FTFT is 20% of students
# interpretation: net_price reflects minority of students
```

### Analyzing Aid Effectiveness

```python
# Equity analysis
gap = net_price_high_income - net_price_low_income

# Positive gap = higher income pays more (progressive)
# Negative gap = lower income pays more (regressive)
# Zero = same net price regardless of income
```

```r
# Equity analysis
gap <- net_price_high_income - net_price_low_income

# Positive gap = higher income pays more (progressive)
# Negative gap = lower income pays more (regressive)
# Zero = same net price regardless of income
```

### Grant Aid Coverage

```python
# What share of costs are covered by grants?
grant_coverage = total_grants / cost_of_attendance * 100

# High coverage = more affordable
# Low coverage = more loans/family contribution needed
```

```r
# What share of costs are covered by grants?
grant_coverage <- total_grants / cost_of_attendance * 100

# High coverage = more affordable
# Low coverage = more loans/family contribution needed
```

## Variable Reference

> **IMPORTANT:** The Portal does NOT use NCES survey variable names (e.g., `scugffn`, `npist1`).
> Instead, the Portal restructures SFA data into **long/tidy format** with dimension columns
> (coded integers) and measure columns (counts, amounts, rates). NCES wide-format variables
> like `npist1`-`npist5` become rows distinguished by `income_level` codes 1-5.
>
> Multiple SFA-related datasets exist in the Portal:
> - `ipeds/colleges_ipeds_sfa_ftft` — FTFT student aid (1999-2017)
> - `ipeds/colleges_ipeds_sfa_grants_and_net_price` — Grants and net price (2008-2021)
> - `ipeds/colleges_ipeds_sfa_all_undergrads` — All undergrad aid (2007-2017)
> - `ipeds/colleges_ipeds_sfa_by_living_arrangement` — Aid by living arrangement (2008-2017)
> - `ipeds/colleges_ipeds_sfa_by_tuition_type` — Aid by tuition type (1999-2017)
>
> Use `get_codebook_url()` from `fetch-patterns.md` to download codebooks:
> ```python
> url = get_codebook_url("ipeds/codebook_colleges_ipeds_sfa_grants_and_net_price")
> ```

### Portal Column Reference by Dataset (Verified)

#### SFA FTFT (`sfa_ftft`) — 12 columns

| Portal Column | Type | Description |
|---------------|------|-------------|
| `unitid` | Int64 | Institution identifier |
| `year` | Int64 | Data year |
| `fips` | Int64 | State FIPS code |
| `type_of_aid` | Int64 | Aid type code (see coded values below) |
| `ftpt` | Int64 | Full-time/part-time status (coded) |
| `level_of_study` | Int64 | Level of study (coded) |
| `class_level` | Int64 | Class level (coded) |
| `degree_seeking` | Int64 | Degree-seeking status (coded) |
| `number_of_students` | Int64 | Count of students receiving this aid type |
| `percent_of_students` | Float64 | Proportion of students receiving aid |
| `average_amount` | Float64 | Average award amount (dollars) |
| `total_amount` | Float64 | Total amount awarded (dollars) |

**`type_of_aid` codes in this dataset:** 1, 2, 3, 4, 5, 6, 7, 8, 10, 11, 12 (code `6` is present in the data — live-verified 2026-08-07; its label is not documented in the codebook excerpt, so verify meaning in the `sfa_ftft` codebook before relying on it)

> **Note on `type_of_aid` codes:** These are integer codes identifying aid categories. Definitions vary by dataset. In `sfa_grants_and_net_price`, code `9` = **all grant and scholarship aid** (Pell + institutional + state/local combined) -- this is NOT Pell-specific. Consult the codebook for the specific dataset you are using: `get_codebook_url("ipeds/codebook_colleges_ipeds_sfa_ftft")`.

#### SFA Grants and Net Price (`sfa_grants_and_net_price`) — 15 columns

| Portal Column | Type | Description |
|---------------|------|-------------|
| `unitid` | Int64 | Institution identifier |
| `year` | Int64 | Data year |
| `fips` | Int64 | State FIPS code |
| `type_of_aid` | Int64 | Aid type code (3 or 9 in this dataset) |
| `income_level` | Int64 | Income bracket code (see below) |
| `ftpt` | Int64 | Full-time/part-time status (coded) |
| `level_of_study` | Int64 | Level of study (coded) |
| `class_level` | Int64 | Class level (coded) |
| `degree_seeking` | Int64 | Degree-seeking status (coded) |
| `tuition_type` | Int64 | Tuition type (coded) |
| `number_of_students` | Int64 | Count of students in group |
| `number_receiving_grants` | Int64 | Count receiving grants |
| `total_grant` | Int64 | Total grant amount (dollars) |
| `average_grant` | Int64 | Average grant amount (dollars) |
| `net_price` | Int64 | Net price (COA minus grants, dollars) |

**`income_level` codes:** 1=$0-$30,000, 2=$30,001-$48,000, 3=$48,001-$75,000, 4=$75,001-$110,000, 5=$110,001+, 99=Total

#### SFA All Undergraduates (`sfa_all_undergrads`) — 10 columns

| Portal Column | Type | Description |
|---------------|------|-------------|
| `unitid` | Int64 | Institution identifier |
| `year` | Int64 | Data year |
| `fips` | Int64 | State FIPS code |
| `type_of_aid` | Int64 | Aid type code (3, 5, or 11 in this dataset) |
| `ftpt` | Int64 | Full-time/part-time status (coded) |
| `level_of_study` | Int64 | Level of study (coded) |
| `number_of_students` | Int64 | Count of students |
| `percent_of_students` | Float64 | Proportion of students |
| `average_amount` | Float64 | Average award amount (dollars) |
| `total_amount` | Float64 | Total amount awarded (dollars) |

#### SFA by Living Arrangement (`sfa_by_living_arrangement`) — 11 columns

| Portal Column | Type | Description |
|---------------|------|-------------|
| `unitid` | Int64 | Institution identifier |
| `year` | Int64 | Data year |
| `fips` | Int64 | State FIPS code |
| `type_of_aid` | Int64 | Aid type code |
| `living_arrangement` | Int64 | Living arrangement code |
| `ftpt` | Int64 | Full-time/part-time status (coded) |
| `level_of_study` | Int64 | Level of study (coded) |
| `class_level` | Int64 | Class level (coded) |
| `degree_seeking` | Int64 | Degree-seeking status (coded) |
| `tuition_type` | Int64 | Tuition type (coded) |
| `number_of_students` | Int64 | Count of students |

#### SFA by Tuition Type (`sfa_by_tuition_type`) — 12 columns

| Portal Column | Type | Description |
|---------------|------|-------------|
| `unitid` | Int64 | Institution identifier |
| `year` | Int64 | Data year |
| `fips` | Int64 | State FIPS code |
| `tuition_type` | Int64 | Tuition type code |
| `type_of_cohort` | Int64 | Cohort type code |
| `ftpt` | Int64 | Full-time/part-time status (coded) |
| `level_of_study` | Int64 | Level of study (coded) |
| `class_level` | Int64 | Class level (coded) |
| `degree_seeking` | Int64 | Degree-seeking status (coded) |
| `number_of_students` | Int64 | Count of students |
| `percent_of_cohort` | Float64 | Proportion of cohort |
| `percent_of_undergrads` | Float64 | Proportion of all undergrads |

### How NCES Variables Map to Portal Structure

The Portal converts NCES wide-format variables into long/tidy rows. For example:

**Student counts** — NCES has separate variables per aid type (`scugpel` for Pell, `scugfsl` for federal loans, etc.). The Portal stores all of these as `number_of_students` rows differentiated by `type_of_aid` code:

| NCES Variable | Portal Equivalent |
|---------------|-------------------|
| `scugffn` | `number_of_students` in `sfa_grants_and_net_price` (filtered by `type_of_aid`) |
| `scugrad` | `number_of_students` in `sfa_ftft` (filtered by `type_of_aid`) |
| `scugpel` | `number_of_students` in `sfa_ftft` where `type_of_aid` = Pell code |
| `scugfsl` | `number_of_students` in `sfa_ftft` where `type_of_aid` = federal loan code |

**Aid amounts** — similarly collapsed:

| NCES Variable | Portal Equivalent |
|---------------|-------------------|
| `upgrnta` | `average_amount` in `sfa_all_undergrads` (filtered by `type_of_aid`) |
| `upgrnt` | `total_amount` in `sfa_all_undergrads` (filtered by `type_of_aid`) |
| `uagrnta` | `average_amount` in `sfa_ftft` (filtered by `type_of_aid`) |
| `uagrntt` | `total_amount` in `sfa_ftft` (filtered by `type_of_aid`) |

**Net price by income** — NCES uses `npist1`-`npist5`; the Portal uses `net_price` column with `income_level` dimension:

| NCES Variable | Portal Equivalent |
|---------------|-------------------|
| `npist1` | `net_price` where `income_level` = 1 ($0-$30K) |
| `npist2` | `net_price` where `income_level` = 2 ($30K-$48K) |
| `npist3` | `net_price` where `income_level` = 3 ($48K-$75K) |
| `npist4` | `net_price` where `income_level` = 4 ($75K-$110K) |
| `npist5` | `net_price` where `income_level` = 5 ($110K+) |
| `npgrn1`-`npgrn5` | `number_of_students` where `income_level` = 1-5 |

### Querying Net Price by Income (Example)

```python
import polars as pl

# Mirror root + pinned revision (keep in sync with education-data-query
# references/mirrors.yaml: HF root_url + vintage.hf_revision). "vintage" = a dated
# Portal snapshot; "revision" = the HF git ref pinning byte-exact bytes.
MIRROR = "https://huggingface.co/datasets/brhkim/education_data_portal_mirror_2026q3/resolve"
REVISION = "0ad00ce0e232c96b0642459e4e7326607a8d26aa"  # immutable commit SHA of the v2 upload
url = f"{MIRROR}/{REVISION}/ipeds/colleges_ipeds_sfa_grants_and_net_price.parquet"
df = pl.read_parquet(url)

# Net price by income level for a specific institution and year
net_prices = (
    df.filter(
        (pl.col("unitid") == 166027)  # Example: MIT
        & (pl.col("year") == 2020)
        & (pl.col("income_level").is_in([1, 2, 3, 4, 5]))
        & (pl.col("type_of_aid") == 9)
    )
    .select("income_level", "net_price", "number_of_students")
    .sort("income_level")
)
```

```r
library(arrow)
library(dplyr)

# Mirror root + pinned revision (keep in sync with education-data-query
# references/mirrors.yaml: HF root_url + vintage.hf_revision). "vintage" = a dated
# Portal snapshot; "revision" = the HF git ref pinning byte-exact bytes.
MIRROR <- "https://huggingface.co/datasets/brhkim/education_data_portal_mirror_2026q3/resolve"
REVISION <- "0ad00ce0e232c96b0642459e4e7326607a8d26aa"  # immutable commit SHA of the v2 upload
url <- paste0(MIRROR, "/", REVISION, "/ipeds/colleges_ipeds_sfa_grants_and_net_price.parquet")
# NOTE: illustrative only — mirror parquet files are Polars-written and may
# declare string_view columns, so a plain read can fail under R arrow
# ("cannot handle Array of type <utf8_view>"). Real fetch scripts must use the
# view-safe parquet read from `education-data-query/references/fetch-patterns.md`.
df <- read_parquet(url)

# Net price by income level for a specific institution and year
net_prices <- df |>
  filter(
    unitid == 166027,
    year == 2020,
    income_level %in% c(1, 2, 3, 4, 5),
    type_of_aid == 9
  ) |>
  select(income_level, net_price, number_of_students) |>
  arrange(income_level)
```

### Cost Variables

Cost of attendance components are in the **tuition/fees datasets**, not SFA datasets:
- `ipeds/colleges_ipeds_ay_tuition_fees` — Academic year tuition and fees
- `ipeds/colleges_ipeds_ay_room_board_other` — Room, board, and other expenses

Consult those codebooks for column names:
```python
url = get_codebook_url("ipeds/codebook_colleges_ipeds_ay_tuition_fees")
```

```r
# get_codebook_url() is a Python helper; in R, build the URL from the mirror root.
# Mirror failover: see `education-data-query/references/fetch-patterns.md` (R pattern)
mirror <- yaml::read_yaml("mirrors.yaml")$mirrors[[1]]
url <- paste0(mirror$root_url, "/", mirror$vintage$hf_revision, "/", "ipeds/codebook_colleges_ipeds_ay_tuition_fees", ".xls")
```

#### NCES Cost Variable Names (for reference only)

These appear in NCES documentation but are NOT used in the Portal:

| NCES Name | Description | Portal Dataset |
|-----------|-------------|----------------|
| `tuition2` | In-state tuition and fees | `ay_tuition_fees` |
| `tuition3` | Out-of-state tuition and fees | `ay_tuition_fees` |
| `roomamt` | Room charges | `ay_room_board_other` |
| `boardamt` | Board charges | `ay_room_board_other` |
| `bksupply` | Books and supplies estimate | `ay_room_board_other` |
| `rmbrdamt` | Room and board combined | `ay_room_board_other` |
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 194kUpdated Claude CodeCodexVerifiedInstall