CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

learn

Interactive lesson tutor for the AI Engineering from Scratch curriculum. Reads LEARNING.md, fetches the next lesson, teaches it section by section in the terminal, quizzes at the end, and records progress. Works cloned or entirely over raw.githubusercontent.com — no setup required. Trigger phrases: "next lesson", "teach me", "continue the course", "let's learn", "resume learning"

Quality

87%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

81%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

An exceptionally well-specified interactive workflow: clear step sequence, complete branch handling, and concrete formats throughout. The main costs are repeated handoff boilerplate, unshown data shapes (quiz.json, a worked lesson path), and inlining sibling-route routing that could be factored into a reference file.

Suggestions

Consolidate the three handoff sections into one parametrized handoff pattern (source-of-truth path plus route-specific gate names) to remove the repeated 'do not choose the next numeric Phase 13 lesson / do not copy state into LEARNING.md' boilerplate.

Add one worked example of a lesson path (e.g. phases/01-llm-fundamentals/what-is-a-token/docs/en.md) and a minimal quiz.json snippet so Step 2/3 handling is fully executable without inspecting the repo first.

Move the resume-routing table and route-owner rules into a references/ file (e.g. ROUTING.md) referenced one level deep, keeping only the dispatch decision inline in SKILL.md.

DimensionReasoningScore

Conciseness

The body is dense and imperative with no concept padding ('Walk equations step by step; ask them to predict the next step'), assuming Claude's competence. Not 5: the three handoff sections repeat near-identical phrasing ('Do not choose the next numeric Phase 13 lesson and do not copy ... into LEARNING.md') that could be consolidated, and a few rationale sentences could be trimmed.

4 / 5

Actionability

Guidance is concrete: exact file paths, the raw.githubusercontent.com URL template, the reply format 'Reply with one letter: <A|B|C|D>.', and the exact LEARNING.md log row contents. Not 5: lesson paths remain as '<phase-dir>/<lesson-dir>' placeholders with no worked example, and the shape of quiz.json is never shown.

4 / 5

Workflow Clarity

Steps 0-5 are clearly sequenced with full decision coverage — resume routing enumerates every branch (named route, single owner, ambiguous, none), Step 0 handles found/complete/missing states. Explicit feedback loops exist: per-answer verdicts, 'Score below 70%: add the lesson to the Review queue', and 'If the learner gets both wrong, offer to re-do that lesson'. Not 4: checkpoints and recovery paths are explicit throughout, not mostly present.

5 / 5

Progressive Disclosure

A single self-contained file, well-sectioned with clear headers and scannable navigation, and remote content sources are precisely specified. Not 5: there are no local reference files at all, and roughly 70 lines of sibling-skill routing rules (resume routing, three handoff sections) are inlined content that could live in a shared reference file.

4 / 5

Total

17

/

20

Passed

Description

88%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: concrete, comprehensive action list with explicit natural trigger phrases and no filler. Its only weakness is that the resume/continue triggers are shared with sibling route skills, creating minor overlap risk.

DimensionReasoningScore

Specificity

The description lists five concrete actions — 'Reads LEARNING.md, fetches the next lesson, teaches it section by section in the terminal, quizzes at the end, and records progress' — which comprehensively covers the skill's workflow. Not 4: coverage is complete rather than having minor gaps.

5 / 5

Completeness

Both 'what' (tutor that reads, teaches, quizzes, and records progress) and 'when' (explicit concrete trigger phrases) are clearly answered. Not 4: the trigger guidance is as explicit and specific as the anchor-5 example, not merely present with room to improve.

5 / 5

Trigger Term Quality

'Trigger phrases: "next lesson", "teach me", "continue the course", "let's learn", "resume learning"' gives multiple natural phrases a user would actually say. Not 5: common variations like 'study', 'lesson time', or the LEARNING.md filename are absent; not 3: several genuinely natural phrasings are present.

4 / 5

Distinctiveness Conflict Risk

The curriculum name, LEARNING.md state file, and distinct trigger phrases give it a clear niche, but 'continue the course' and 'resume learning' overlap with the sibling routes (learn-mcp, learn-agent-skills, learn-reverse-engineering) the skill itself routes to. Not 5: that shared vocabulary creates minor conflict risk with closely related skills.

4 / 5

Total

18

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

rohitg00/ai-engineering-from-scratch

Path

skills/learn/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.