DAAF-Contribution-Community/daaf / .claude/skills/education-data-source-crdc
education-data-source-crdc
VerifiedCRDC — OCR civil rights collection for U.S. public schools; Portal families have topic-specific coverage, with year labels from 2011 through 2023 (Portal year = fall of the school year, so 2020 = 2020-21). Discipline, course access, harassment, restraint/seclusion by race/sex/disability/EL. Use for civil rights and equity analysis. OCR documents every collection since 2011-12 as a universe collection; 2020-21 (Portal year 2020) is COVID-impacted.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 26.5 KB |
| references/civil-rights-context.md | markdown | 15.6 KB |
| references/collection-methodology.md | markdown | 16.4 KB |
| references/data-elements.md | markdown | 24.2 KB |
| references/data-quality.md | markdown | 22.8 KB |
| references/historical-changes.md | markdown | 16.7 KB |
| references/variable-definitions.md | markdown | 22.7 KB |
references/data-elements.md
View on GitHub# CRDC Data Elements
Comprehensive reference for all data categories collected in the Civil Rights Data Collection. This document covers what is collected, how it's disaggregated, and analytical considerations.
## Contents
- [Enrollment](#enrollment)
- [Discipline](#discipline)
- [Restraint and Seclusion](#restraint-and-seclusion)
- [Harassment and Bullying](#harassment-and-bullying)
- [Chronic Absenteeism](#chronic-absenteeism)
- [Course Access and Offerings](#course-access-and-offerings)
- [Advanced Courses (AP/IB/Gifted)](#advanced-courses-apibgifted)
- [Students with Disabilities](#students-with-disabilities)
- [English Learners](#english-learners)
- [Preschool](#preschool)
- [Retention](#retention)
- [Staffing](#staffing)
- [School Offenses](#school-offenses)
- [School Characteristics](#school-characteristics)
---
## Enrollment
### What's Collected
Total student enrollment disaggregated by:
- Race/ethnicity (7 categories)
- Sex (male/female)
- Disability status (IDEA-served, 504-only)
- English learner status
### Key Variables
| Portal Column | Description |
|---------------|-------------|
| `enrollment_crdc` | Enrollment count (varies by row's race/sex/disability/lep combination) |
| `psenrollment_crdc` | Preschool enrollment count |
> **Portal Data Structure:** Unlike OCR raw files which have separate columns per subgroup (e.g., `enrollment_black`, `enrollment_male`), the Portal uses a row-based structure. Each row represents a unique combination of school + race + sex + disability + lep. To get total enrollment for a school, filter to `race == 99, sex == 99, disability == 99, lep == 99`. To get Black enrollment, filter to `race == 2, sex == 99, disability == 99, lep == 99`.
### Analytical Notes
- `enrollment_crdc` serves as **denominator** for rate calculations
- Match enrollment year with outcome year carefully
- Some variables use different enrollment bases (e.g., discipline uses school enrollment, AP uses high school enrollment)
- The `psenrollment_crdc` column contains many `-2` (not applicable) values for non-preschool schools
### Cross-Tabulations Available
| Level | Available |
|-------|-----------|
| Race only | Yes |
| Sex only | Yes |
| Race × Sex | Yes |
| Disability only | Yes |
| Race × Disability | Limited |
| Sex × Disability | Limited |
---
## Discipline
### What's Collected
School discipline data is the most extensive CRDC category, covering multiple discipline types for both K-12 and preschool students.
### Discipline Categories
| Category | Definition |
|----------|------------|
| **In-school suspension (ISS)** | Student removed from regular class but remains at school |
| **Out-of-school suspension (OSS)** | Student removed from school for disciplinary reasons |
| **One suspension** | Students receiving exactly one suspension |
| **More than one suspension** | Students receiving multiple suspensions |
| **Expulsion with services** | Removed from school, educational services provided |
| **Expulsion without services** | Removed from school, no educational services |
| **Referral to law enforcement** | Any referral to police/law enforcement |
| **School-related arrest** | Arrest occurring on school grounds or for school incident |
| **Corporal punishment** | Physical discipline (permitted in some states) |
### Discipline Variables
| Portal Column Name | Description |
|-------------------|-------------|
| `students_susp_in_sch` | In-school suspension count |
| `students_susp_out_sch_single` | One out-of-school suspension |
| `students_susp_out_sch_multiple` | More than one OSS |
| `expulsions_with_ed_serv` | Expulsion with educational services |
| `expulsions_no_ed_serv` | Expulsion without services |
| `expulsions_zero_tolerance` | Zero-tolerance expulsions |
| `students_referred_law_enforce` | Referrals to law enforcement |
| `students_arrested` | School-related arrests |
| `students_corporal_punish` | Corporal punishment |
| `transfers_alt_sch_disc` | Transfers to alternative school for discipline |
> **Note:** The Portal uses a row-based structure where each row represents a combination of school + race + sex + disability + LEP. Subgroup breakdowns are obtained by filtering on the categorical columns (e.g., `race == 2` for Black students), not by separate variable names. The variable names above are the actual Portal column names as observed in `schools_crdc_discipline_k12` data.
### Disaggregation
| Disaggregation | Available |
|----------------|-----------|
| Race/ethnicity | Yes (7 categories) |
| Sex | Yes |
| Disability | Yes (IDEA students) |
| Race × Sex | Yes |
| Race × Disability | Yes |
| Sex × Disability | Yes |
### Duration Information (2015-16+)
These items were optional in 2013-14 and required from 2015-16 (Portal: 2015+, in the `discipline-instances` and `suspensions-days` families):
- **Instances of suspension** - Total suspensions (students can have multiple)
- **School days missed** - Total days of instruction missed due to OSS
### Calculating Discipline Rates
```python
import polars as pl
def discipline_rate_by_group(df, discipline_var, enrollment_var='enrollment_crdc', per=100):
"""
Calculate discipline rate per 100 (or 1000) students.
Args:
df: DataFrame filtered to a subgroup (e.g., race==2 for Black)
discipline_var: Column with discipline counts
enrollment_var: Column with enrollment counts
per: Rate multiplier (100 or 1000)
Returns:
Rate per specified multiplier
"""
# Filter out coded missing values
df_valid = df.filter(pl.col(discipline_var) >= 0)
total_disciplined = df_valid.select(pl.col(discipline_var).sum()).item()
total_enrolled = df_valid.select(pl.col(enrollment_var).sum()).item()
if total_enrolled == 0:
return None
return (total_disciplined / total_enrolled) * per
# Example: OSS rate for Black students (race=2 in Portal encoding)
black_students = df.filter(pl.col('race') == 2)
black_oss_rate = discipline_rate_by_group(
black_students,
'students_susp_out_sch_single',
per=100
)
```
```r
library(dplyr)
# Calculate discipline rate per 100 (or 1000) students.
# Example: OSS rate for Black students (race=2 in Portal encoding)
discipline_var <- "students_susp_out_sch_single"
enrollment_var <- "enrollment_crdc"
per <- 100 # rate multiplier (100 or 1000)
# Subgroup of interest: Black students
df_group <- df |> filter(race == 2)
# Filter out coded missing values (-1/-2/-3)
df_valid <- df_group |> filter(.data[[discipline_var]] >= 0)
total_disciplined <- sum(df_valid[[discipline_var]], na.rm = TRUE)
total_enrolled <- sum(df_valid[[enrollment_var]], na.rm = TRUE)
black_oss_rate <- if (total_enrolled > 0) {
total_disciplined / total_enrolled * per
} else {
NA_real_
}
```
### Important Caveats
1. **Definition of "suspension" varies** - Length thresholds differ by district
2. **Informal discipline not captured** - Hallway talks, unofficial removals not counted
3. **Policy changes affect data** - Zero-tolerance policies inflate counts
4. **COVID impact** - 2020-21 (Portal year 2020) shows artificially low discipline due to remote learning
5. **Underreporting likely** - Schools may underreport sensitive discipline data
---
## Restraint and Seclusion
### What's Collected
Data on physical intervention practices with students, particularly those with disabilities.
### Categories
| Category | Definition |
|----------|------------|
| **Mechanical restraint** | Use of device to restrict movement |
| **Physical restraint** | Personal physical contact to restrict movement |
| **Seclusion** | Involuntary confinement in room/area student cannot leave |
### Variables
| Variable | Description |
|----------|-------------|
| `students_restrained_mechanical` | Students subjected to mechanical restraint |
| `students_restrained_physical` | Students subjected to physical restraint |
| `students_secluded` | Students subjected to seclusion |
| `instances_restraint_mechanical` | Total instances of mechanical restraint |
| `instances_restraint_physical` | Total instances of physical restraint |
| `instances_seclusion` | Total instances of seclusion |
### Disaggregation
| Disaggregation | Students | Instances |
|----------------|----------|-----------|
| Total | Yes | Yes |
| Race/ethnicity | Yes | No |
| Sex | Yes | No |
| Disability (IDEA) | Yes | Yes |
### Key Concerns
- **Disproportionate use with students with disabilities** - Most restraint/seclusion involves IDEA students
- **Definition variation** - States and districts define differently
- **Underreporting** - Documentation practices inconsistent
- **Deaths and injuries** - Serious incidents have occurred
### Policy Context
OCR has issued guidance that restraint and seclusion should:
- Only be used when necessary for safety
- Never be used as punishment
- Be documented and reviewed
- Trigger review of student's behavioral plan
---
## Harassment and Bullying
### What's Collected
Reports of harassment or bullying, collected by type of harassment and by whether students were disciplined.
### Harassment Categories
| Category | Definition |
|----------|------------|
| **Based on race, color, national origin** | Targeting student's race, ethnicity, or national origin |
| **Based on sex** | Sexual harassment, gender-based harassment |
| **Based on disability** | Targeting student's disability |
| **Based on sexual orientation** | Targeting perceived sexual orientation |
| **Based on religion** | Targeting student's religion |
### Variables
For each category, CRDC collects:
| Variable Type | Description |
|---------------|-------------|
| `allegations_{type}` | Students alleging harassment |
| `disciplined_{type}` | Students disciplined for harassment |
### Disaggregation
| Variable | Race | Sex | Disability |
|----------|------|-----|------------|
| Allegations | Yes | Yes | Yes |
| Disciplined | Yes | Yes | Yes |
### Important Distinctions
- **Allegation** - Student (or parent) reported harassment
- **Discipline** - Student was disciplined for harassing behavior
- These are **different students** - victims vs. perpetrators
### Analytical Cautions
1. **Reporting vs. occurrence** - Low allegations may mean underreporting, not less harassment
2. **Definition variation** - What constitutes "harassment" varies
3. **Procedure differences** - Schools have different reporting mechanisms
4. **Substantiation not captured** - CRDC doesn't distinguish substantiated vs. unsubstantiated
---
## Chronic Absenteeism
### What's Collected (2013-14+)
Students chronically absent from school, defined as **missing 15 or more school days** during the school year.
### Variable
| Variable | Description |
|----------|-------------|
| `chronic_absent` | Students missing 15+ days |
### Disaggregation
- Race/ethnicity (7 categories)
- Sex
- Disability status (IDEA)
- English learner status
### Key Considerations
1. **Definition** - The CRDC form used 15+ school days (2013-14, 2015-16). From 2017-18 the data come from EDFacts (state-reported); a uniform 15-day definition across states after 2016 was not verified.
2. **Any absence counts** - Excused and unexcused absences
3. **COVID impact** - 2020-21 values come from EDFacts (state-reported) and reflect pandemic attendance disruption; compare with caution
4. **Source change** - First collected in the 2013-14 CRDC and collected again in 2015-16. From 2017-18 on, OCR takes chronic-absenteeism data from EDFacts (state-reported) rather than the CRDC form; treat the 2015 to 2017 change as a source change
### Chronic Absenteeism vs. Truancy
| Measure | Definition |
|---------|------------|
| **Chronic absenteeism** | Missing 15+ days (any reason) |
| **Truancy** | Unexcused absences only |
CRDC collects chronic absenteeism, not truancy.
---
## Course Access and Offerings
### What's Collected
Information on courses **offered** by schools and student **enrollment** in specific courses.
### Mathematics Courses
| Course | Description |
|--------|-------------|
| Algebra I | First-year algebra |
| Algebra II | Second-year algebra |
| Geometry | Geometry |
| Advanced mathematics | Courses beyond Algebra II |
| Calculus | Calculus |
### Science Courses
| Course | Description |
|--------|-------------|
| Biology | Biology |
| Chemistry | Chemistry |
| Physics | Physics |
### Computer Science (2017-18+)
| Variable | Description |
|----------|-------------|
| `offers_cs` | School offers any computer science |
| `cs_enrollment` | Students enrolled in CS |
### Disaggregation
| Level | Available |
|-------|-----------|
| School offers course | Yes (school-level indicator) |
| Student enrollment | By race/ethnicity, sex |
### Equity Analysis
```python
import polars as pl
# Example: AP course access disparity
def course_access_analysis(df):
"""
Analyze whether schools serving more minority students
are less likely to offer advanced courses.
Uses Portal integer codes: race=2 (Black), race=3 (Hispanic)
"""
# Get enrollment totals by race for each school
# Note: race=99 is total enrollment in Portal data
totals = df.filter(pl.col('race') == 99).select(['ncessch', 'enrollment_crdc'])
black = df.filter(pl.col('race') == 2).select(['ncessch', pl.col('enrollment_crdc').alias('enrollment_black')])
hispanic = df.filter(pl.col('race') == 3).select(['ncessch', pl.col('enrollment_crdc').alias('enrollment_hispanic')])
school_df = totals.join(black, on='ncessch', how='left').join(hispanic, on='ncessch', how='left')
school_df = school_df.with_columns(
((pl.col('enrollment_black').fill_null(0) + pl.col('enrollment_hispanic').fill_null(0)) /
pl.col('enrollment_crdc')).alias('pct_minority')
)
# Group by minority percentage quartile
school_df = school_df.with_columns(
pl.when(pl.col('pct_minority') <= 0.25).then(pl.lit('0-25%'))
.when(pl.col('pct_minority') <= 0.50).then(pl.lit('25-50%'))
.when(pl.col('pct_minority') <= 0.75).then(pl.lit('50-75%'))
.otherwise(pl.lit('75-100%'))
.alias('minority_quartile')
)
return school_df
```
```r
library(dplyr)
library(tidyr)
# Example: AP course access disparity
# Analyze whether schools serving more minority students
# are less likely to offer advanced courses.
# Uses Portal integer codes: race=2 (Black), race=3 (Hispanic)
# Get enrollment totals by race for each school
# Note: race=99 is total enrollment in Portal data
totals <- df |> filter(race == 99) |> select(ncessch, enrollment_crdc)
black <- df |> filter(race == 2) |> select(ncessch, enrollment_black = enrollment_crdc)
hispanic <- df |> filter(race == 3) |> select(ncessch, enrollment_hispanic = enrollment_crdc)
school_df <- totals |>
left_join(black, by = "ncessch") |>
left_join(hispanic, by = "ncessch")
school_df <- school_df |> mutate(
pct_minority = (replace_na(enrollment_black, 0) + replace_na(enrollment_hispanic, 0)) /
enrollment_crdc
)
# Group by minority percentage quartile
school_df <- school_df |> mutate(
minority_quartile = case_when(
pct_minority <= 0.25 ~ "0-25%",
pct_minority <= 0.50 ~ "25-50%",
pct_minority <= 0.75 ~ "50-75%",
.default = "75-100%"
)
)
```
---
## Advanced Courses (AP/IB/Gifted)
### What's Collected
Enrollment in advanced academic programs designed to provide rigorous coursework.
### Categories
| Program | Description |
|---------|-------------|
| **AP (Advanced Placement)** | College-level courses with standardized exams |
| **IB (International Baccalaureate)** | International college-prep curriculum |
| **Gifted and Talented** | Programs for academically gifted students |
### AP Course Categories
| Category | Courses Included |
|----------|-----------------|
| Mathematics | AP Calculus AB/BC, AP Statistics |
| Science | AP Biology, Chemistry, Physics, Environmental Science |
| Computer Science | AP Computer Science A, Principles |
| Other AP | All other AP courses |
### Variables (Portal Column Names)
| Variable | Description |
|----------|-------------|
| `enrl_ap` | Total AP enrollment |
| `enrl_ap_math` | AP math enrollment |
| `enrl_ap_science` | AP science enrollment |
| `enrl_ap_compsci` | AP computer science enrollment |
| `enrl_ap_language` | AP language enrollment |
| `enrl_ap_other` | AP other enrollment |
| `enrl_ib` | Students enrolled in IB program |
| `enrl_gifted_talented` | Students in gifted/talented |
> **Note:** These column names are empirically confirmed from the `schools_crdc_apib_enroll` dataset. Each row represents a combination of school + race + sex + disability + lep. Filter categorical columns to get subgroup-specific enrollments.
### Disaggregation
| Program | Race | Sex | Disability |
|---------|------|-----|------------|
| AP enrollment | Yes | Yes | No |
| IB enrollment | Yes | Yes | No |
| Gifted/Talented | Yes | Yes | Yes |
### Key Equity Concerns
1. **Access disparities** - Not all schools offer AP/IB
2. **Enrollment gaps** - Even when offered, enrollment differs by race/sex
3. **Gifted identification** - Disproportionate identification by race
4. **Resource inequity** - Funding differences for programs
---
## Students with Disabilities
### What's Collected
Information about students served under IDEA and Section 504.
### Categories
| Category | Definition |
|----------|------------|
| **IDEA students** | Students with IEPs under IDEA |
| **Section 504 only** | Students with 504 plans (not IEP) |
### IDEA Disability Categories
| Category | Code |
|----------|------|
| Autism | AUT |
| Deaf-blindness | DB |
| Emotional disturbance | ED |
| Hearing impairment | HI |
| Intellectual disability | ID |
| Multiple disabilities | MD |
| Orthopedic impairment | OI |
| Other health impairment | OHI |
| Specific learning disability | SLD |
| Speech or language impairment | SLI |
| Traumatic brain injury | TBI |
| Visual impairment | VI |
| Developmental delay | DD |
### Educational Settings
| Setting | Description |
|---------|-------------|
| Inside regular class 80%+ | Most time in general education |
| Inside regular class 40-79% | Mixed placement |
| Inside regular class <40% | Mostly separate setting |
| Separate school/facility | Entirely separate placement |
### Variables
| Variable | Description |
|----------|-------------|
| `enrollment_idea` | Total IDEA students |
| `enrollment_504_only` | 504 plan students (not IDEA) |
| `idea_setting_{type}` | IDEA students by educational setting |
| `idea_category_{type}` | IDEA students by disability category |
---
## English Learners
### What's Collected
Information about students classified as English learners (EL) / Limited English Proficient (LEP).
### Variables
| Variable | Description |
|----------|-------------|
| `enrollment_lep` | English learner students |
| `el_programs` | Types of EL programs offered |
### Program Types (collected variably)
- Dual language immersion
- Transitional bilingual
- English as a Second Language (ESL)
- Newcomer programs
### Key Considerations
1. **Definition variation** - EL classification criteria vary by state
2. **Exit criteria** - When students are reclassified varies
3. **Title VI implications** - Failure to serve ELs may violate civil rights
---
## Preschool
### What's Collected
Enrollment and discipline data for preschool (ages 3-5, pre-kindergarten).
### Preschool Discipline Variables
| Variable | Description |
|----------|-------------|
| `pk_oss_one` | Preschoolers with one OSS |
| `pk_oss_more` | Preschoolers with multiple OSS |
| `pk_expulsion` | Preschoolers expelled |
### Disaggregation
- Race/ethnicity
- Sex
- Disability status
### Major Concern
**Preschool suspension is a significant civil rights issue**:
- Very young children suspended at high rates
- Significant racial disparities
- Questions about developmental appropriateness
### Research Finding
The 2011-12 CRDC (Data Snapshot: School Discipline, March 2014) found Black children were 18% of preschool enrollment but 48% of preschoolers receiving more than one OSS; in 2013-14 they were 19% of enrollment and 47% of preschoolers receiving one or more OSS.
---
## Retention
### What's Collected
Students retained in grade (held back to repeat a grade).
### Variable
| Variable | Description |
|----------|-------------|
| `retained` | Students retained in grade |
### Disaggregation
- Race/ethnicity
- Sex
- Disability status (IDEA)
### Analytical Note
Retention rates vary significantly by:
- State policy (social promotion policies)
- Grade level (transition grades often higher)
- Local policy changes
---
## Staffing
### What's Collected
Information about school staff, including teachers, counselors, and support personnel.
### Staff Categories
| Category | Description |
|----------|-------------|
| **Teachers** | Classroom teachers |
| **Counselors** | School counselors |
| **Psychologists** | School psychologists |
| **Social workers** | School social workers |
| **Nurses** | School nurses |
| **Security/SROs** | Security guards, School Resource Officers |
### Teacher Variables
| Variable | Description |
|----------|-------------|
| `teachers_fte` | Full-time equivalent teachers |
| `teachers_first_year` | First-year teachers |
| `teachers_second_year` | Second-year teachers |
| `teachers_certified` | Fully certified teachers |
| `teachers_not_certified` | Teachers not fully certified |
### Teacher Experience Variables
| Variable | Description |
|----------|-------------|
| `teachers_chronic_absent` | Teachers chronically absent (15+ days) |
| `teachers_salary_avg` | Average teacher salary |
### Support Staff
| Variable | Description |
|----------|-------------|
| `counselors_fte` | School counselor FTE |
| `psychologists_fte` | School psychologist FTE |
| `social_workers_fte` | School social worker FTE |
| `nurses_fte` | School nurse FTE |
### Equity Analysis
Staffing data enables analysis of:
- Teacher experience gaps by school demographics
- Counselor ratios by school characteristics
- SRO presence and discipline correlation
---
## School Offenses
### What's Collected
Incidents occurring at school, regardless of whether they resulted in discipline.
### Offense Categories
| Category | Description |
|----------|-------------|
| **Violent incidents without weapons** | Fights, assaults without weapons |
| **Violent incidents with weapons** | Assaults involving weapons |
| **Firearm possession** | Possession of firearm at school |
| **Other weapons possession** | Possession of knife, other weapon |
| **Drug-related incidents** | Drug possession, distribution, use |
| **Alcohol-related incidents** | Alcohol possession or use |
### Variables
| Variable | Description |
|----------|-------------|
| `offenses_violent_no_weapon` | Violent incidents without weapons |
| `offenses_violent_weapon` | Violent incidents with weapons |
| `offenses_firearm` | Firearm incidents |
| `offenses_weapon_other` | Other weapon incidents |
| `offenses_drug` | Drug incidents |
| `offenses_alcohol` | Alcohol incidents |
### Distinctions
- **Offense** - Incident occurred
- **Discipline** - Student was disciplined (different data element)
- **Referral/arrest** - Law enforcement involved (different data element)
---
## School Characteristics
### What's Collected
Basic information about school characteristics.
### Variables
| Variable | Description |
|----------|-------------|
| `grades_offered` | Grade levels served (PS-12) |
| `magnet_school` | Magnet school indicator |
| `charter_school` | Charter school indicator |
| `alternative_school` | Alternative school indicator |
| `virtual_school` | Virtual/online school |
| `single_sex_school` | Single-sex school |
| `single_sex_classes` | Offers single-sex classes |
### Note on Linking
School characteristics should be linked from **CCD** (Common Core of Data) for:
- Urbanicity
- School level
- Total enrollment
- Free/reduced lunch eligibility
- Location (state, district)
---
## Data Element Changes Over Time
| Data Element | First Collected | Notes |
|--------------|-----------------|-------|
| Basic discipline | 2011 (modern) | |
| Restraint/seclusion | 2011 | |
| Chronic absenteeism | 2013 | Collected again 2015; from EDFacts 2017+ (source change) |
| Preschool discipline | 2011 | Expanded 2015 |
| Sexual harassment | 2011 | |
| Computer science | 2017 | |
| Instances of suspension | 2013 | Optional in 2013-14; required from 2015-16 (Portal: 2015+) |
| School days missed | 2013 | Optional in 2013-14; required from 2015-16 (Portal: 2015+) |
| COVID-related items | 2020 | Temporary |
See `./historical-changes.md` for detailed year-by-year changes.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |