AIDRIN¶

Domain: AI data readiness, aidrin metrics (quality, fairness, privacy) on UCI Adult

Source: use_cases/aidrin-ai-readiness/

Apply AIDRIN through DSAgt to a single tabular dataset (UCI Adult), 15 metrics spanning data-quality, impact-on-AI, fairness-and-bias, and data-governance.

Estimated time: ~15 minutes

This tutorial demonstrates AIDRIN (AI Data Readiness Inspector) on a single tabular dataset: metrics from all four of AIDRIN's categories, with execution provenance. For how the aidrin skill and code are set up and what rules the agent follows, see the AI-Readiness Check page.

The dataset is the UCI Adult census extract included with AIDRIN (examples/sample_data/csv/adult.csv). The demonstrated AIDRIN metrics rely on: a record ID, quasi-identifiers (age, sex, race), sensitive attributes (sex, race), and a prediction target (income).

Applied Metrics¶

Category	Metrics
data-quality	`completeness`, `duplicity`, `outliers`
impact-of-data-on-AI	`correlations`, `feature-relevance`
fairness-and-bias	`class-imbalance`, `statistical-rates`, `representation-rate`
data-governance	`k-anonymity`, `l-diversity`, `t-closeness`, `entropy-risk`, `single-attribute-risk`, `multiple-attribute-risk`, `differential-privacy`

Prerequisites¶

DSAgt installed and an agent platform installed and already authenticated
Python 3.12 or later

Setup¶

dsagt init

At the menu, name the project aidrin, pick your agent, and keep the defaults. Then fetch the sample dataset into the project and start the session:

PROJ=~/dsagt-projects/aidrin
mkdir -p "$PROJ/data"
curl -sL https://raw.githubusercontent.com/idtlab/AIDRIN/v2026.08.2/examples/sample_data/csv/adult.csv \
    -o "$PROJ/data/adult.csv"
dsagt start aidrin

Execution¶

Paste these prompts one at a time.

1. Confirm the AIDRIN skill is installed¶

Using the aidrin skill, list the readiness metrics AIDRIN provides.

Verify: the agent reads skills/aidrin/SKILL.md and its reference/metrics.md and lists the metrics by category; it may also run aidrin list through dsagt-run.

2. Run the metrics¶

Using the aidrin skill, run a readiness assessment on data/adult.csv. Cover these
four categories:
(1) data-quality: completeness, duplicity, outliers;
(2) impact-of-data-on-AI: correlations on "age,education.num,sex,race", and feature-relevance with
    categorical columns "workclass,education,sex,race", numerical columns
    "age,education.num,hours.per.week", target income;
(3) fairness-and-bias: class-imbalance on income, statistical-rates on income with sensitive
    attribute sex, representation-rate on "sex,race";
(4) data-governance: k-anonymity on "age,sex,race", l-diversity on "age,sex,race" with sensitive
    column income, t-closeness on "age,sex,race" with sensitive column income, entropy-risk on
    "age,sex,race", single-attribute-risk with id-column ID and eval-columns "age,sex,race",
    multiple-attribute-risk with id-column ID and eval-columns "age,sex,race", and
    differential-privacy on "age,hours.per.week" with epsilon 1.0.
Then give me a readiness verdict organized by the four categories.

Expect: the exact commands and representative results (positional args; JSON to stdout):

Data quality

The skill runs the three quality metrics as one baseline call.

Command	Result
`aidrin data-quality data/adult.csv --detail`	completeness `1.0`; duplicity `0.0`; outliers overall `≈0.050` (`hours.per.week` ≈0.277)

Impact on AI

Command	Result
`aidrin run correlations data/adult.csv "age,education.num,sex,race"`	Theil's U + Pearson matrices
`aidrin run feature-relevance data/adult.csv "workclass,education,sex,race" "age,education.num,hours.per.week" income`	Pearson-to-target (e.g. `education.num` ≈0.34, `age` ≈0.23)

Fairness & bias

Command	Result
`aidrin run class-imbalance data/adult.csv income`	imbalance degree `≈0.52`
`aidrin run statistical-rates data/adult.csv income sex`	Female `>50K` ≈11% vs Male ≈31%
`aidrin run representation-rate data/adult.csv "sex,race"`	Male:Female ≈2.0, White:Black ≈8.9

Data governance / privacy

Command	Result
`aidrin run k-anonymity data/adult.csv "age,sex,race"`	`k = 1`
`aidrin run l-diversity data/adult.csv "age,sex,race" income`	`l = 1`
`aidrin run t-closeness data/adult.csv "age,sex,race" income`	`t ≈ 0.76`
`aidrin run entropy-risk data/adult.csv "age,sex,race"`	`≈0.06`
`aidrin run single-attribute-risk data/adult.csv ID "age,sex,race"`	per-attribute risk stats
`aidrin run multiple-attribute-risk data/adult.csv ID "age,sex,race"`	joint re-identification risk
`aidrin run differential-privacy data/adult.csv "age,hours.per.week" 1.0`	noised mean/variance per column; also writes `noisy/noisy_data.csv`

The agent should produce a four-part verdict: quality is clean (complete, no duplicates, moderate hours.per.week outliers); impact shows education.num/age as the strongest income predictors; fairness flags a large gender gap in the target (men ~2.8× more likely >50K); and governance flags severe re-identification risk (k = 1, l = 1) on the age,sex,race quasi-identifiers; bin or suppress before sharing.

3. Batch several metrics from one config¶

Write an aidrin batch config (YAML) that runs completeness, class-imbalance, statistical-rates, and
representation-rate on data/adult.csv with target income and sensitive attribute sex, then run it
with the aidrin skill. The config is one flat mapping, and the key names are in the skill's
reference/metrics.md.

The config for this step:

file_path: data/adult.csv
metrics: [completeness, class-imbalance, statistical-rates, representation-rate]
target-column: income
y-true-column: income
sensitive-attribute-column: sex
columns: [sex, race]

4. Generate a datacard from the assessment¶

Use the datacard-generator skill to write a Level 1 datacard for data/adult.csv that incorporates
the readiness findings above. Take the values from the dataset and the reports, note anything
unknown rather than asking, and write it as one file, data/genesis_datacard_adult.md.

The agent follows the datacard-generator base skill: it fills the skill's template from the dataset and the readiness reports, writes one Genesis Datacard, data/genesis_datacard_adult.md, and validates it with the registered datacard-validate code. The validator warns that the filename differs from the one it derives from the dataset name; that warning is expected, since the prompt fixes the filename.

5. Review the execution records¶

Show me the execution records for this session as a table of metric, command, and exit code.

The agent reads the records dsagt-run wrote to trace_archive/ and lists one row per aidrin command: the runs from step 2 (thirteen with the quality baseline as one data-quality call, fifteen when the agent runs the three quality metrics separately) and the batch run from step 3, every exit code 0.

6. Review the project artifacts¶

Show me the contents of my project folder in a tree format, with the artifacts dsagt recorded during this session highlighted. Include the registered codes and installed skills.

Expect: a listing of the whole project directory, including the registered codes and installed skills under skills/, with a line on what each entry is. The listing marks the execution records in trace_archive/ (each aidrin record holds that metric's report), the datacard, the trace store mlflow.db, and the session's other outputs.

Post-Conditions¶

skills/aidrin/SKILL.md is present, with a PROVENANCE.txt naming the AIDRIN source.
trace_archive/ holds one execution record per aidrin command from step 2, at least thirteen.
Results span the four categories, with the gender-fairness gap and the k = 1 / l = 1 re-identification risks identified.
One datacard for the dataset exists, data/genesis_datacard_adult.md.
The agent lists every metric call from the execution records with its command and exit code.
MLflow traces capture token usage, latency, and the code-execution spans.

Coverage¶

DSAgt Capability	Steps
The `aidrin` base skill and code installed at init	Setup
Base-skill use: the `aidrin` CLI through `dsagt-run`	1
Code execution with provenance (execution records in `trace_archive/`)	2
Multi-metric orchestration	2
Multi-metric / batch execution	3
Base-skill use (`datacard-generator`)	4
Provenance review from the execution records	5
Review of the session's artifacts	6
Observability (MLflow spans in the serverless `mlflow.db` store)	all

View the traces any time with dsagt traces aidrin.

Cleanup¶

dsagt rm aidrin -y