recolligere

Read durable disk records to reconstitute context after memory erasure.

Updated May 20, 2026
One-click install
npx skills add https://github.com/watmin/datamancy.dev --skill recolligere
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: recolligere
Source: https://github.com/watmin/datamancy.dev/tree/main/recolligere
Command: npx skills add https://github.com/watmin/datamancy.dev --skill recolligere

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Recolligere is a primer to reconstitute orientation after memory erasure by reading the durable record on disk rather than relying on the live cache, preventing fluent but hollow self-contradictions and ensuring a grounded starting point.

Core Features & Use Cases

  • Read the local recovery map and the state-of-world artifacts before proposing work.
  • Ground actions in the persisted record rather than memory, reducing drift after resets.

Quick Start

Fetch the local recovery file, read the live state from disk, and resume from the recovered orientation.

Frequently Asked Questions about recolligere

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

FAQPage Schema
How do I restore project context after an AI model session reset?

Restore project context after a session reset by reading the durable record on disk rather than the cached memory. Fetch a local recovery file to map workspace boundaries, read the live breadcrumb to identify the current state, and reconstitute orientation before acting.

Why does my model contradict itself after memory erasure despite sounding fluent?

Models contradict themselves after memory erasure because they rely on cached memory instead of a durable record. Reading the state-of-world artifact from disk grounds the model's decisions and prevents the fluent but hollow self-contradictions that occur without a stable orientation.

What files do I need to set up before recovering orientation from disk?

Recovering orientation from disk requires a local recovery file to map workspace boundaries, a live breadcrumb to identify the current state, and a state-of-world artifact as the authoritative source for decisions. These persisted records must exist before the session starts.

How do I resume work from a recovered orientation after a reset?

Resume work from a recovered orientation by fetching the local recovery file, reading the live state from disk, and grounding your next actions in the persisted record. This ensures you resume from the recovered context rather than proposing work from a hollow memory.

What is the best way to prevent context drift across multiple AI project resets?

Prevent context drift across project resets by grounding actions in a durable record on disk rather than live cache. Reconstituting a stable orientation from a state-of-world artifact ensures continuity matters are respected and decisions remain consistent across sessions.

When should I not rely on cached memory for AI session continuity?

Do not rely on cached memory for session continuity after memory erasure or post-reset sessions. Cached memory causes drift and self-contradictions; always reconstitute orientation by reading the authoritative state-of-world artifact and live breadcrumb from disk instead.