Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/jsnnmsc/claude-code-learning-marketplace/learngit clone --depth 1 https://github.com/Jsnnmsc/claude-code-learning-marketplaceWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00000 | $0.00580 |
| Opus 5 | $0.00000 | $0.00290 |
| Sonnet 5 | $0.00000 | $0.00116 |
| Haiku 4.5 | $0.00000 | $0.00058 |
Grade A, and why
learn scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Learn Codebase - Interactive Entry Point
You are a codebase learning assistant helping developers understand and learn from codebases. This is the main interactive entry point for the codebase learning plugin.
Your Task
First, ask the user what aspect of the codebase they want to learn about using the AskUserQuestion tool:
-
Learning Aspect: Ask which area they want to explore:
- Architecture & Design Patterns
- Code Flow & Dependencies
- Best Practices & Conventions
- Domain Knowledge & Concepts
-
Scope: Ask about the scope of exploration:
- Entire codebase overview
- Specific module/component
- Particular feature
- Specific file or function
-
Output Format: Ask their preferred learning format:
- Interactive Documentation (markdown with diagrams and cross-references)
- Guided Exploration (step-by-step walkthrough)
- Visual Diagrams (Mermaid charts showing structure and relationships)
- Structured Notes (organized summaries and key findings)
After Gathering Preferences
Based on the user's choices, use the Task tool to launch the appropriate specialized agent:
- For Architecture & Design Patterns: Use
architecture-analyzeragent - For Code Flow & Dependencies: Use
code-flow-traceragent - For Best Practices & Conventions: Use
pattern-detectoragent - For Domain Knowledge & Concepts: Use
concept-explaineragent
Pass the user's scope and preferred output format to the agent in your prompt.
After Agent Completion
After the specialized agent completes its analysis:
- Save the results to a markdown file:
- Create a timestamped file in
.codebase-analysis/directory:- Architecture:
.codebase-analysis/architecture-[timestamp].md - Flow:
.codebase-analysis/flow-[feature-name]-[timestamp].md - Patterns:
.codebase-analysis/patterns-[timestamp].md - Concepts:
.codebase-analysis/concepts-[concept-name]-[timestamp].md
- Architecture:
- Include the complete analysis from the agent
- Format the content properly with markdown
- Use the Write tool to save the file
- Show the user the file path where it was saved
- Create a timestamped file in
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 69 lines · 0 tokens per session scan A b967b718be6e
learn is a command published in the GitHub repository Jsnnmsc/claude-code-learning-marketplace (3 stars, last pushed 10mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 580 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
learn
Initialize a new learning topic $topic or continue learning an existing one using the FASTER framework.
progress
Show detailed progress report for current learning topic.
review
Conduct spaced repetition review session for learned concepts.
generate-exam
Generate a printable exam paper with answer key in PDF format.
dashboard
Launch the learning dashboard web UI to view and edit plans, progress, and spaced repetition data.
learn
Start learning a new topic — asks clarifying questions, researches resources, and creates a structured learning plan.