dec-bench-snapshot

Convert DEC Bench run sets into Linear snapshots with parent issues and documents.

1|2|Updated Feb 22, 2026
One-click install
npx skills add https://github.com/514-labs/agent-evals --skill dec-bench-snapshot
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: dec-bench-snapshot
Source: https://github.com/514-labs/agent-evals/tree/main/.claude/skills/dec-bench-snapshot
Command: npx skills add https://github.com/514-labs/agent-evals --skill dec-bench-snapshot

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Capture and preserve completed DEC Bench evaluations by converting a run set into a durable Linear snapshot, including a parent issue, a Linear Document, and labeled follow-up sub-issues.

Core Features & Use Cases

  • Automates the end-to-end creation of a snapshot workflow: parent AXP issue, companion Document, and linked follow-up tasks, all wired to a specific run set.
  • Manages prerequisites and governance, including gcloud authentication checks and explicit labeling to enable reproducible replay and auditability.
  • Supports multi-scenario, multi-harness DEC Bench runs with trace uploads to GCS to ensure traceable, reviewable results.

Quick Start

Ask me to snapshot the latest DEC Bench run set and I will generate the Linear parent issue, create a companion document, and seed follow-up issues with proper labels.

Frequently Asked Questions about dec-bench-snapshot

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

FAQPage Schema
How do I automate capturing DEC Bench evaluation runs as Linear issues?▼

Automating DEC Bench snapshots converts a run set into a Linear parent issue, a companion Document, and labeled follow-up sub-issues to preserve and organize evaluation results.

What is a Linear snapshot for multi-scenario evaluation runs?▼

A Linear snapshot packages DEC Bench evaluation runs by creating a parent AXP issue, a companion document for step-by-step artifact packaging, and linked follow-up tasks for pivoting to work.

Do I need gcloud authentication to create DEC Bench snapshots?▼

Yes, creating DEC Bench snapshots requires gcloud authentication checks as a prerequisite to enable GCS trace uploads, ensuring reproducible replay and auditability for the evaluation runs.

How do I upload DEC Bench traces to GCS for issue tracking?▼

Uploading DEC Bench traces to GCS is managed during the snapshot workflow, connecting the traceable results to the generated Linear issues and companion document for reviewable auditability.

Can I use this snapshot process for multi-harness DEC Bench runs?▼

Yes, the snapshot workflow supports multi-harness and multi-scenario DEC Bench runs, applying explicit labeling and an AGENTS.md delegation contract to enforce governance across the evaluation set.