Review, search, prune, and export project learnings from JSONL files.
npx skills add https://github.com/a29paul/hexlens --skill learn-a29paul
Please help me install this Agent Skill. Skill: learn Source: https://github.com/a29paul/hexlens/tree/main/.claude/skills/gstack/learn Command: npx skills add https://github.com/a29paul/hexlens --skill learn-a29paul
Project insights and past solutions get lost across chat sessions and scattered notes, forcing teams to repeatedly solve the same problems. This skill centralizes everything the AI has learned about your project into a searchable, maintainable knowledge base.
Use the learn skill to review recent project learnings and search for past patterns related to the current branch.
High-intent search queries and answers about installing and using this skill.
To recover project knowledge across sessions, you need a context recovery tool that performs semantic search on local learnings JSONL files. This allows you to instantly recall past patterns, review recent insights grouped by type, and avoid repeating solved problems.
You can search for past development patterns using natural language queries through semantic search functionality. This mechanism scans all recorded project learnings to find specific insights, pitfalls, or solutions without requiring exact keyword matches.
To detect stale project knowledge referencing deleted files, use staleness pruning functionality. This mechanism automatically flags outdated learnings that reference deleted files or contradict newer entries, keeping your institutional knowledge base maintainable.
Yes, you can export project learnings for external sharing. The export feature extracts your centralized development insights from the local knowledge base, allowing you to distribute institutional knowledge and view statistics about captured sessions.
Yes, you need access to local gstack analytics directories and learnings JSONL files to review captured insights. This environment setup is required to perform semantic search, staleness detection, and cross-session context recovery.
The best way to maintain institutional knowledge in AI-assisted environments is centralizing insights into a searchable knowledge base. This approach surfaces relevant past learnings proactively when similar problems occur, preventing knowledge loss across scattered chat sessions.
limzivectorManage and retrieve learnings from a gstack project session JSONL file.
shekerkammaManages cross-session project learnings, insights, and pitfalls captured by gstack.
ShmurdocCapture and organize project learnings into a searchable JSONL knowledge base.
GlourOrganize and persist cross-session project knowledge in a ByteRover context tree.
