agentsclimarketplace

Fyuu tutor

Skill fyuuwang/fyuu-tutor/fyuu-tutor

Document-backed adaptive tutoring for capability, certification, and language learning

Install
npx -y skills add fyuuwang/fyuu-tutor --skill fyuu-tutor

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 21 days oldThe repository was created 21 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Operate long-running, document-backed tutoring projects that turn trusted materials or goals into adaptive HTML lessons, practice, and evidence-guided next steps. Use when creating, continuing, reviewing, or handing off a private learning project; auditing source materials; planning a curriculum; generating lessons or reference aids; recording learner evidence; or routing a project through capability, certification, or language tutoring.

SKILL.md

5.9 KB, ~1.2k tokens by cl100k_base, as published. Nobody here has run it

Fyuu Tutor

Treat project files as the control plane. Keep reusable tutoring logic in this Skill and all learner data, sources, generated content, and live state in the private project.

Route the project

Choose the primary pipeline by the evidence that will prove success:

  • Capability: the learner must explain, apply, build, decide, or transfer. Read capability.md.
  • Certification: the learner must satisfy a dated exam outline and perform under exam conditions. Read certification.md.
  • Language: the learner must understand or produce a target language. Read language.md.

If the success evidence is ambiguous, ask one question that would change the route. Do not route by input format: a PDF, topic, case, or question bank can support any suitable pipeline.

Start

  1. Locate the project and require project.toml plus pipeline.toml.
  2. Run python3 scripts/validate_project.py --project <project-dir>.
  3. Read core.md and project-schema.md.
  4. Read the selected pipeline above. When materials must be acquired or converted, also read material-pipeline.md.
  5. Read teaching-loop.md before planning a lesson, evaluating an attempt, or choosing the next step.
  6. Before producing any lesson HTML, read ui-contract.md and confirm the UI kit is installed.
  7. Read the declared private profile, then MISSION.md, NOTES.md, CURRICULUM.md when present, and STATUS.md.

Operate

  1. Confirm synchronization and require STATUS.md to be idle.
  2. Claim one bounded project task with update_status.py before changing project artifacts.
  3. Preserve source files and keep production, study, demonstrated performance, and stable mastery distinct.
  4. Apply the shared material and teaching loops plus only the selected pipeline.
  5. Put generated lessons, aids, evidence, and history only in the project paths declared by project.toml.
  6. Run deterministic checks and record remaining manual uncertainty.
  7. If the private workspace has a documented standing portal authorization and its private portal config enables incremental publishing, run the portal wrapper after releasing the claim. Before production, check the public Skill repository is clean on main; a dirty public-source worktree is a blocker, never a reason to stash or commit unrelated changes.

Pipeline defaults yield to the private learner profile and project rules for teaching choices. Nothing may override the safety and state rules in core.md.

Finish

  1. Run the project validator, the smallest pipeline-specific check, and check_links.py.
  2. Record produced artifacts, learner evidence, unresolved errors, blockers, and one next action separately.
  3. Rebuild indexes after changing HTML output or STATUS.md.
  4. Release the claim and verify STATUS.md is idle.

Deterministic tools

  • Project lifecycle: scripts/create_project.py, validate_project.py, update_status.py.
  • Output checks: scripts/check_links.py, build_index.py, audit_privacy.py.
  • Material processing: scripts/pdf/.
  • Optional pipeline implementations: scripts/pipelines/.
  • UI kit: scripts/sync_ui_kit.py (install/check/upgrade) and scripts/validate_lesson_ui.py (--file/--project/--kit/--self-test).
  • Learning portal: scripts/build_portal.py (explicit --project only; never publishes all projects by default). Deployment via scripts/deploy_portal.sh requires --publish and is only called through the wrapper.
  • Offline export: scripts/export_offline_lesson.py creates a self-contained single-file HTML by inlining local lesson.css and lesson.js. Use it when a reader needs a standalone copy for Safari or offline reading. It rejects remote resources, file:// links, and missing assets and never modifies the source file.
  • Portal publish wrapper: scripts/publish_portal.py reads a private <workspace>/portal.toml to determine which projects are eligible for publishing, runs all validators, builds the portal, scans for private data, and calls deploy_portal.sh --publish. publish_after_validation is an enforced technical eligibility gate and defaults to false when absent; it is not authorization. A workspace-level standing authorization requires an automatic --publish after a covered course is produced and released; current-session authority is for a one-off release. Do not ask again for ordinary courses covered by standing authorization. A dirty public Skill worktree, a new project/source/purpose, a changed portal roster/destination, or any privacy/validation failure is a blocker requiring separate authority. If publishing fails, record the failure reason in STATUS.md without changing course status. Never bypass the wrapper to copy private directories into gh-pages.
  • Public routes are assigned by build_portal.py from ASCII project IDs, never from private folder names or localized display names. Every publishable HTML filename must be lowercase ASCII kebab-case (for example 0002-shared-vision.html); the builder rejects anything else. Keep source-page links local (../assets/, ../index.html); the builder rewrites its published copy to the short public route. Course-generation Agents must not hard-code GitHub Pages URLs.

Do not install dependencies, publish content, expose private sources, or resolve another agent's claim without explicit user authority.

Gives 0 of the 12 instructions most learn study skills give in ~1.2k tokens

Counted across 546 of the 573 authors here whose files we hold, read 2026-08-07

  • calculate the zone of proximal development before teachingin 25 of 546, across 8 files
  • produce self-contained HTML lessonsin 24 of 546, across 8 files
  • record user preferences in a notes filein 23 of 546, across 5 files
  • maintain a teaching workspace in the current directoryin 21 of 546, across 4 files
  • find high-quality resources before writing lessonsin 19 of 546, across 5 files
  • make lessons beautiful, short, and quickly completablein 19 of 546, across 3 files
  • create reusable components for lessonsin 19 of 546, across 5 files
  • create compressed reference documents for quick lookupin 19 of 546, across 3 files
  • update the mission file and records upon mission changesin 16 of 546, across 2 files
  • set min_dist to 0.0 for clustering preprocessingin 16 of 546, across 6 files
  • populate the mission file before teachingin 15 of 546, across 1 file
  • include interactive feedback loops in lessonsin 15 of 546, across 1 file

Said here and by no other author read

  • store learner data and state only in private project files
  • choose the pipeline based on required success evidence
  • require project.toml and pipeline.toml to start
  • read core.md and project-schema.md before operating
  • read teaching-loop.md before planning or evaluating lessons
  • confirm UI kit is installed before producing lesson HTML

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.