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 skills/melsayedx/learning-mode/learning-modenpx skills add melsayedx/learning-mode --skill learning-modegit clone --depth 1 https://github.com/melsayedx/learning-modeWhat 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 | $0.00116 | $0.03191 |
| Opus 5 | $0.00058 | $0.01596 |
| Sonnet 5 | $0.00023 | $0.00638 |
| Haiku 4.5 | $0.00012 | $0.00319 |
Grade A, and why
learning-mode 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 yesterday.
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 — 375 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Learning Mode
Help the learner become able to produce, explain, and transfer the target skill without the assistant doing the load-bearing work for them.
Use agent-neutral language and capabilities. Do not assume a particular host, model, tool name, project-instruction filename, plugin system, or configuration directory.
Operating flow
Follow this sequence:
Determine intent: quick answer, guided learning, or delivery
↓
Define one learning outcome
↓
Check the smallest relevant prerequisite
↓
Select the next missing concept
↓
Predict → attempt → feedback → retry
↓
Validate against the defined outcome
├─ Fails → diagnose, revise, and retry
└─ Passes → continue
↓
Can the learner explain and transfer it?
├─ No → hint ladder or analogous example
└─ Yes → next checkpoint or finish
↓
Update journal only if previously enabled
Do not turn the sequence into a large upfront questionnaire or syllabus. Ask only what is needed for the next useful step.
Activation
Determine the user's intent before applying the teaching constraints:
- Quick answer: answer directly and compactly. Do not force an exercise or create a journal.
- Guided learning: apply this skill.
- Delivery: build, fix, or explain the requested result normally; do not apply this skill.
Treat explicit requests such as "teach me," "guide me," "I want to practice," or "help me build this myself" as guided-learning intent. A one-off "what," "how," or "why" question alone is not enough.
If intent is genuinely ambiguous, ask exactly one question:
Do you want a quick answer, guided learning where you do the core work, or the finished result delivered?
Starting point
Define one observable outcome for the current session. Prefer "implement and explain a safe webhook handler" over "learn webhooks."
Check only the smallest prerequisite that could block that outcome. Use one prediction, explanation, trace, or tiny task. Reuse demonstrated background and any enabled journal; do not make the learner re-prove settled knowledge.
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- yesterday First seen · 375 lines · 116 tokens per session scan A 4201a1df898a
learning-mode is a skill published in the GitHub repository melsayedx/learning-mode (5 stars, last pushed 1mo ago), licensed MIT. It adds 116 tokens to every session and 3,191 once invoked, about $0.0006 per session on Opus 5. 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 skills, from other repositories
init-edu
Use when starting a new educational project or study folder — sets up CLAUDE.md with teaching rules, configures Explanatory output style, initializes the global education DB, and sets up project-local tracking.
demo
Use when the concept needs ANIMATION or INTERACTIVITY to be understood — step-by-step protocol flows, sorting algorithm traces, state machine transitions, network packet journeys, or anything where static images fall short. Generates a single .html file with Canvas/SVG + JavaScript that opens in any browser; styling…
excalidraw
Middle tier — complex STATIC diagrams. Use when user invokes /excalidraw, OR when concept has 5+ components, multi-layer architecture, or detailed data structures that benefit from editable layout. Generates .excalidraw JSON files opened in excalidraw.com or VS Code. For simple visuals use /ascii. For animated/dynamic…
roadmap
Use when the student wants a visual learning roadmap for a goal or topic area — reads current knowledge state, researches the standard learning path, and generates an interactive .excalidraw diagram with color-coded topic nodes, prerequisite arrows, and milestone markers. Visual companion to /research.
flashcards
Use when the student wants to generate Anki-style flashcards from studied topics — creates question/answer cards in Markdown and Anki-compatible CSV, prioritized by weak areas and misconceptions. Cards saved to global education DB and project-local docs.
quiz-me
Use when the user wants to be quizzed, tested, or asked questions on any topic — programming concepts, protocols, algorithms, tools, or any subject they are studying.