StamKavid/last-ds-mile
A guided data-science lifecycle for Claude Code: frame, explore, baseline, validate, model, evaluate, and report — with leakage and honesty checks built into every stage.
ARCHETYPE
Skill Collector
Ten-plus skills loaded on demand. A procedural-knowledge library.
Copy this rig
# review before running: this installs third-party code $ npx degit StamKavid/last-ds-mile/agents ./rig-last-ds-mile/agents $ npx degit StamKavid/last-ds-mile/commands ./rig-last-ds-mile/commands $ npx degit StamKavid/last-ds-mile/skills ./rig-last-ds-mile/skills $ npx degit StamKavid/last-ds-mile/hooks ./rig-last-ds-mile/hooks
MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit StamKavid/last-ds-mile/skills/capturing-learnings .claude/skills/capturing-learnings $ npx degit StamKavid/last-ds-mile/skills/causal-vs-predictive .claude/skills/causal-vs-predictive $ npx degit StamKavid/last-ds-mile/skills/data-science-project .claude/skills/data-science-project $ npx degit StamKavid/last-ds-mile/skills/data-viz-standards .claude/skills/data-viz-standards $ npx degit StamKavid/last-ds-mile/skills/dataframe-performance .claude/skills/dataframe-performance $ npx degit StamKavid/last-ds-mile/skills/distribution-shift .claude/skills/distribution-shift $ npx degit StamKavid/last-ds-mile/skills/ds-baseline .claude/skills/ds-baseline $ npx degit StamKavid/last-ds-mile/skills/ds-brief .claude/skills/ds-brief $ npx degit StamKavid/last-ds-mile/skills/ds-data .claude/skills/ds-data $ npx degit StamKavid/last-ds-mile/skills/ds-deploy .claude/skills/ds-deploy $ npx degit StamKavid/last-ds-mile/skills/ds-evaluate .claude/skills/ds-evaluate $ npx degit StamKavid/last-ds-mile/skills/ds-explain .claude/skills/ds-explain $ npx degit StamKavid/last-ds-mile/skills/ds-explore .claude/skills/ds-explore $ npx degit StamKavid/last-ds-mile/skills/ds-frame .claude/skills/ds-frame $ npx degit StamKavid/last-ds-mile/skills/ds-handoff .claude/skills/ds-handoff $ npx degit StamKavid/last-ds-mile/skills/ds-iterate .claude/skills/ds-iterate $ npx degit StamKavid/last-ds-mile/skills/ds-method .claude/skills/ds-method $ npx degit StamKavid/last-ds-mile/skills/ds-model .claude/skills/ds-model $ npx degit StamKavid/last-ds-mile/skills/ds-package .claude/skills/ds-package $ npx degit StamKavid/last-ds-mile/skills/ds-prep .claude/skills/ds-prep $ npx degit StamKavid/last-ds-mile/skills/ds-report .claude/skills/ds-report $ npx degit StamKavid/last-ds-mile/skills/ds-validate .claude/skills/ds-validate $ npx degit StamKavid/last-ds-mile/skills/error-analysis .claude/skills/error-analysis $ npx degit StamKavid/last-ds-mile/skills/imbalanced-data .claude/skills/imbalanced-data $ npx degit StamKavid/last-ds-mile/skills/metric-selection .claude/skills/metric-selection $ npx degit StamKavid/last-ds-mile/skills/model-ensembling .claude/skills/model-ensembling $ npx degit StamKavid/last-ds-mile/skills/notebook-hygiene .claude/skills/notebook-hygiene $ npx degit StamKavid/last-ds-mile/skills/target-leakage-detection .claude/skills/target-leakage-detection $ npx degit StamKavid/last-ds-mile/skills/uncertainty-quantification .claude/skills/uncertainty-quantification
$ curl -fsSL --create-dirs -o .claude/agents/data-profiler.md https://raw.githubusercontent.com/StamKavid/last-ds-mile/main/agents/data-profiler.md $ curl -fsSL --create-dirs -o .claude/agents/ds-reviewer.md https://raw.githubusercontent.com/StamKavid/last-ds-mile/main/agents/ds-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/leakage-auditor.md https://raw.githubusercontent.com/StamKavid/last-ds-mile/main/agents/leakage-auditor.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} Skills (29)
capturing-learningscausal-vs-predictivedata-science-projectdata-viz-standardsdataframe-performancedistribution-shiftds-baselineds-briefds-datads-deployds-evaluateds-explainds-exploreds-frameds-handoffds-iterateds-methodds-modelds-packageds-prepds-reportds-validateerror-analysisimbalanced-datametric-selectionmodel-ensemblingnotebook-hygienetarget-leakage-detectionuncertainty-quantification
Subagents (3)
| data-profiler model: haiku | Fast structural profiling sweep for a dataset — shape, dtypes, missingness, cardinality, duplicate keys. Use during /ds-data or /ds-explore for a quick first-pass profile. Not for deep statistical ana |
| ds-reviewer model: sonnet | Runs the ds-method discipline checklist against a notebook or pipeline before /ds-report — baseline present, validation strategy sound, slice performance checked, metric matches the problem. Use befor |
| leakage-auditor model: opus | Adversarially hunts for target leakage across a feature pipeline — features that encode the target directly, temporal leakage where future information reaches training data, and validation-split leaka |
Slash commands (17)
/ds-baseline/ds-brief/ds-data/ds-deploy/ds-evaluate/ds-explain/ds-explore/ds-frame/ds-handoff/ds-iterate/ds-learn/ds-model/ds-package/ds-prep/ds-report/ds-validate/ds
Similar rigs
Expeed-Software/exloom
Spec-driven development workflow for Claude Code, built for teams — with an enforced, evidence-based review gate. MIT-licensed.
Skill Collector 857 tok · · Q0
alexandre-darmon/ai-pm-skills
Evals as evidence: tested Agent Skills for AI product managers in regulated companies
Skill Collector 3.2k tok · · Q0
mayankmankhand/llm-peer-review
Multi-model AI debate for your entire project lifecycle. Plans, specs, research, and code. Free and open source.
Orchestrator 4.0k tok · · Q15
OpenLAIR/dr-claw
A Super AI Lab with massive AI Doctors as Assistants. Best IDE for Research via AI Power.
Skill Collector 8.2k tok · · Q15