Persist project learnings to .agents/LEARNINGS.md and consult at session start.
npx skills add https://github.com/abpai/skills --skill project-memory-abpai
Please help me install this Agent Skill. Skill: project-memory Source: https://github.com/abpai/skills/tree/main/project-memory Command: npx skills add https://github.com/abpai/skills --skill project-memory-abpai
Keeps a persistent, per-project memory that captures learnings, mistakes, and preferences to prevent repeating errors and accelerate future work.
.agents/LEARNINGS.md after each task.Read or create .agents/LEARNINGS.md at the start of each session and begin logging corrections, patterns, and preferences as you work.
High-intent search queries and answers about installing and using this skill.
You maintain project memory by logging corrections, patterns, and preferences to a LEARNINGS.md file after each task, then reading it at session start to prime decisions. This prevents repeating errors and accelerates future work.
A living project memory file captures learnings, mistakes, and preferences persistently throughout a project. It speeds up development by letting sessions read past context to avoid repeating errors and accelerate future work.
You log learnings by writing corrections and patterns to the .agents/LEARNINGS.md file after each task. The template is created automatically if missing, ensuring continuous knowledge retention without manual setup.
You need a root SKILL.md file with YAML frontmatter containing a name and description. The learnings file resides in the .agents directory, and you can use optional scripts, references, and assets directories to extend functionality.
The best way to manage recurring insights is through periodic maintenance that merges and summarizes patterns in your LEARNINGS.md file. This keeps long-term knowledge retention organized and ensures session priming remains efficient.
SteveBronderWrite or update agent memory files with YAML frontmatter.
jackchukaAggregate work activities from multiple sources into a reflective journal with improvements.
gancimOrganizes knowledge into PARA-based memory files with YAML frontmatter for cross-session recall.
4lermanSave knowledge entries to auto-memory with timestamp and context.