datagen-reduce-dataset-snapshots

Reduce Daytona snapshot counts by grouping Dockerfiles into shared environments.

276|37|Updated Dec 5, 2025

Search Tags:#reduce #daytona #dockerfiles #dataset-snapshots #patchers #oracle-validation

Ready to run

npx skills add https://github.com/open-thoughts/OpenThoughts-Agent --skill datagen-reduce-dataset-snapshots

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: datagen-reduce-dataset-snapshots
Source: https://github.com/open-thoughts/OpenThoughts-Agent/tree/main/.claude/skills/datagen-reduce-dataset-snapshots
Command: npx skills add https://github.com/open-thoughts/OpenThoughts-Agent --skill datagen-reduce-dataset-snapshots

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Reduce Daytona's unique-environment snapshot counts for Harbor datasets by adjusting patcher environment-build logic to group Dockerfiles and reuse environments where possible.

Core Features & Use Cases

Group per-task Dockerfiles into shared environment images to dramatically cut snapshots.
Diagnose env-hash driver issues, regenerate + upload, and re-count with a two-tier quality gate (infra smoke + oracle) to validate reductions.
Real-world example: SweGym dataset reduction from 906 to 5 snapshots demonstrates the payoff of grouping.

Quick Start

Execute the patcher to group Dockerfiles by shared environments and re-count snapshots locally to confirm the cap is under 10.

Frequently Asked Questions about datagen-reduce-dataset-snapshots

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I reduce Docker environment snapshots in Daytona datasets?▼

To reduce unique-environment snapshots in Daytona, group per-task Dockerfiles into shared environment images by adjusting patcher environment-build logic to reuse environments where possible.

What causes SnapshotCapExceeded errors in Harbor task datasets?▼

SnapshotCapExceeded errors occur when Harbor task datasets generate too many unique environments. Diagnosing the env-hash driver and grouping Dockerfiles before regenerating snapshots addresses this cap.

How do I validate dataset quality after reducing snapshots?▼

Validate snapshot reductions using a two-tier quality gate that runs an infrastructure smoke test followed by oracle validation to ensure the grouping did not hurt dataset quality.

Can I run snapshot reduction locally on Mac with Daytona?▼

Yes, the snapshot reduction workflow runs deterministically on Mac plus Daytona, requiring you to document thresholds, regenerations, and oracle validation locally to confirm compliance.

How many snapshots can I cut by grouping Dockerfiles?▼

Grouping Dockerfiles can dramatically cut snapshots; a real-world SweGym dataset example reduced unique environments from 906 snapshots down to just 5.

What is the best way to diagnose env-hash driver issues for dataset snapshots?▼

Diagnose env-hash driver issues by adjusting the patcher environment-build logic to group Dockerfiles, then regenerate, upload, and re-count snapshots locally to confirm the cap is under 10.