dr-next-snapshot-replay

Resume staged tasks from validated snapshot files with a bilingual replay prompt.

11|1|Updated Apr 10, 2026
One-click install
npx skills add https://github.com/Arcanada-one/datarim --skill dr-next-snapshot-replay
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dr-next-snapshot-replay
Source: https://github.com/Arcanada-one/datarim/tree/main/skills/dr-next-snapshot-replay
Command: npx skills add https://github.com/Arcanada-one/datarim --skill dr-next-snapshot-replay

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents redoing already-completed work by restarting the next stage from a validated snapshot, ensuring consistent continuity across task resumptions.

Core Features & Use Cases

  • Snapshot-first resume contract: Reads datarim/snapshots/{TASK-ID}.snapshot.md before any other task inputs for /dr-next and /dr-orchestrate.
  • Deterministic validation with silent fallback: Validates snapshot frontmatter via check-stage-snapshot-on-exit.sh; if missing or malformed, it falls back to legacy behavior without warning noise.
  • Canonical replay prompt emission: Emits a required bilingual replay prompt containing recommended_next and the literal snapshot body to drive autonomous completion.

Quick Start

Use dr-next or dr-orchestrate for a task ID that already has a valid snapshot file at datarim/snapshots/{TASK-ID}.snapshot.md so the system replays from the saved stage context.

Frequently Asked Questions about dr-next-snapshot-replay

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

FAQPage Schema
How do I resume an AI agent pipeline task without re-reading the entire prompt?

To resume a task without re-reading the entire prompt, use a snapshot replay mechanism that reads a validated stage snapshot file containing the previously completed work and recommended next steps. This allows the agent pipeline to continue directly from the saved context.

What is a snapshot-first replay contract in task resumption?

A snapshot-first replay contract is a workflow validation mechanism that prioritizes reading a validated stage snapshot file before any other task inputs. It uses the snapshot's recommended next steps and literal body to drive autonomous task completion without losing prior progress.

How do I validate a stage snapshot before resuming an agent workflow?

To validate a stage snapshot, apply a configured check script to the snapshot file's frontmatter before execution. If the frontmatter is missing or malformed, the system deterministically falls back silently to the legacy read order without generating warning noise.

Can I use snapshot replay to skip previously completed workflow stages?

Yes, snapshot replay enables you to skip completed workflow stages by resuming from a validated snapshot file located at a specific snapshot path. The system emits a canonical bilingual replay prompt using the snapshot's recommended next steps to continue the task.

What happens if a task snapshot file is missing during workflow resumption?

If a task snapshot file is missing during workflow resumption, the system silently falls back to the legacy behavior. It defaults to the standard read order for task descriptions and planning inputs instead of halting the agent pipeline with errors.

Do I need a specific file format to enable task resumption for my AI orchestration workflow?

Yes, task resumption requires a markdown snapshot file containing specific frontmatter and a literal snapshot body. This file must be placed at the designated snapshot directory path and validated by the configured exit check script to ensure prompt contract compliance.