ss-sdd-maintaining-memory-file

Update a project's agent memory file from SDD pipeline artifacts.

Updated May 19, 2026
One-click install
npx skills add https://github.com/Emrebener/Sublime-Skills --skill ss-sdd-maintaining-memory-file
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ss-sdd-maintaining-memory-file
Source: https://github.com/Emrebener/Sublime-Skills/tree/main/skills/spec-driven-development/ss-sdd-maintaining-memory-file
Command: npx skills add https://github.com/Emrebener/Sublime-Skills --skill ss-sdd-maintaining-memory-file

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents stale or bloated agent memory files in a Spec-Driven Development (SDD) pipeline by updating only the project-level truths that matter to future agents.

Core Features & Use Cases

  • Reads SDD artifacts: Ingests the current run’s spec, plan, and ADRs to determine whether any project-level conventions or constraints changed.
  • Earns-it-place updates only: Updates the configured agent memory file only for significant architectural changes or non-obvious project conventions, avoiding narrative or transient content.
  • Respects safety and size limits: Enforces a configurable character cap and refuses to operate when no memory file path is configured or when the file is missing on disk.

Quick Start

Use this skill when dispatched as a subagent during the memory-file maintenance stage of an SDD pipeline run to produce an incremental, budgeted update to your project memory file.

Frequently Asked Questions about ss-sdd-maintaining-memory-file

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

FAQPage Schema
How do I prevent my agent memory file from becoming stale or bloated?

To prevent stale or bloated agent memory, update the memory file only with significant project-level truth changes from spec, plan, and ADR outputs while enforcing a strict character cap.

What is the best way to update project conventions in an SDD pipeline?

The best way to update project conventions in an SDD pipeline is dispatching a subagent during the memory maintenance stage to ingest artifacts and apply earned-its-place updates to the memory file.

How do I manage architectural decision records within an agent's persistent memory?

To manage architectural decision records in agent memory, ingest ADR outputs during the post-implementation stage and add or edit rules only for significant architectural changes, avoiding transient content.

When should I not write to an agent memory file during a pipeline run?

You should not write to an agent memory file when no memory file path is configured or when the target file is missing on disk, as the update process requires strict path validation before writing.

Does the agent memory maintenance process enforce a character budget?

Yes, the agent memory maintenance process enforces a configurable character budget to prevent bloating, ensuring only significant project-level truths are kept accurate and lean.

What is the purpose of a memory maintenance stage in spec-driven development?

The purpose of the memory maintenance stage in spec-driven development is to produce incremental, budgeted updates to the project memory file based on spec, plan, and ADR outputs before pipeline completion.