Data-Wise · Data-Wise · Jun 13, 2026 · Jun 12, 2026 · Jun 12, 2026 · Jun 12, 2026
diff --git a/.STATUS b/.STATUS
@@ -1,9 +1,9 @@
 status: Active
 progress: 100
-next: Post-v2.18.1 — rich-exam v1 grading-line gap (backlog #13); rich-exam v2 examark QTI (backlog #11)
-target: v2.18.1 — Maintenance (dep CVE patch→0 vulns, ESLint gate+cleanup, \printanswers short-answer fix, timing de-flake)
-milestone: v2.18.1 release via PR (dev→main) 2026-06-12 | 33 commands | 38 tutorials | 3,405 tests (3,268 Jest + 137 node:test) | 143 suites
-last_session: 2026-06-12 - spec-status cleanup + PR #107 review/merge + v2.18.1 release pipeline (Session 92)
+next: Homebrew tap auto-update VALIDATED (App 2874502 + HOMEBREW_TAP_APP_ID/PRIVATE_KEY secrets set; workflow_dispatch smoke test green, App-authored tap PR #118 merged) — only residual: confirm de750f5 forwarding rides dev→main on the NEXT real release (release events run from main's copy); then rich-exam v1 grading-line gap (backlog #13); rich-exam v2 examark QTI (backlog #11)
+target: v2.19.0 — scaffolded-math-tutorial teaching skill (18th) + skills surfaced in the hub (/scholar:hub skills, discoverSkills) + skill_count macro + df-conflation closure-audit guardrail; Missing-Effect adversarial-review release (#2)
+milestone: v2.19.0 release via PR (dev→main) 2026-06-13 | 33 commands | 18 skills | 38 tutorials | 3,411 tests (3,268 Jest + 143 node:test) | 144 suites
+last_session: 2026-06-12 (Session 94) - Homebrew tap-auth ACTIVATED + validated end-to-end: created GitHub App "Data-Wise Homebrew Automation" (ID 2874502), set HOMEBREW_TAP_APP_ID + HOMEBREW_TAP_APP_PRIVATE_KEY secrets, proved via workflow_dispatch smoke test (App-authored tap PR #118 merged — also fixed stale manifest.json scholar 2.17.0→2.18.1). Debug gotcha: "JWT could not be decoded" = wrong/revoked private key (secret had old 2026-05-12 .pem; valid was 2026-06-12) — isolated with a local openssl+curl /app test. Also: root-caused `brew upgrade scholar` block to upstream Homebrew 6.0.1-28 build-sandbox regression (commit 080ffb3d1b "deny all of $HOME"), shipped tap bin.mkpath fix (PR #117, all 8 source-built formulae) + scholar upgraded to 2.18.1 via brew (pin-to-6.0.1 workaround, then unpinned). Prior: Session 93 post-v2.18.1 cleanup (de750f5, 4dd2afc, spec v5)
 complexity: Low
 risk_level: Low
 dependencies: Claude API, js-yaml, ajv, ajv-keywords, ajv-formats, Quarto, flow-cli, R/Rscript (runtime), examark (Canvas QTI), jstat (F-distribution)

diff --git a/.claude-plugin/plugin.json b/.claude-plugin/plugin.json
@@ -1,6 +1,6 @@
 {
   "name": "scholar",
-  "version": "2.18.1",
+  "version": "2.19.0",
   "description": "Academic workflows for research and teaching - literature management, manuscript writing, simulation design, and course material generation",
   "author": {
     "name": "Data-Wise"

diff --git a/.github/markdown-link-check-config.json b/.github/markdown-link-check-config.json
@@ -114,6 +114,12 @@
     {
       "pattern": "^https://aspredicted.org"
     },
+    {
+      "pattern": "^https://www.imsglobal.org"
+    },
+    {
+      "pattern": "^https://opensource.org/licenses"
+    },
     {
       "pattern": "%7B%7B"
     },

diff --git a/.github/workflows/homebrew-release.yml b/.github/workflows/homebrew-release.yml
@@ -76,4 +76,10 @@ jobs:
       command_count: '33'
       skill_count: '17'
     secrets:
+      # Preferred: GitHub App token (no expiry). The reusable workflow uses the
+      # App path when both app_id and app_private_key are non-empty, else falls
+      # back to tap_token (a PAT, which expires). Add HOMEBREW_TAP_APP_ID and
+      # HOMEBREW_TAP_APP_PRIVATE_KEY repo secrets to activate the App path.
+      app_id: ${{ secrets.HOMEBREW_TAP_APP_ID }}
+      app_private_key: ${{ secrets.HOMEBREW_TAP_APP_PRIVATE_KEY }}
       tap_token: ${{ secrets.HOMEBREW_TAP_GITHUB_TOKEN }}
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -5,6 +5,23 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [2.19.0] — 2026-06-13
+
+### Added
+
+- **`scaffolded-math-tutorial` teaching skill** — a reusable methodology for turning a paper, proof, or formula into a college-algebra-floor self-study course (four pillars, the Terminology Closure Audit, self-explanation + faded worked examples, dual HTML/PDF Quarto). Brings the skill count to **18** under a new **Teaching** category.
+- **Skills surfaced in the hub** — `/scholar:hub skills` lists auto-activating skills by category, and the overview gains a SKILLS block. The discovery engine adds `discoverSkills()` / `getSkillStats()` / `getSkillsByCategory()` (scanning `skills/**/SKILL.md`), with 7 new node:test cases — zero changes to the command-scan path or cache.
+- **`skill_count` macro** in `mkdocs.yml`, synced by `scripts/version-sync.js` and consumed by `docs/skills.md`.
+- `docs/WHATS-NEW-v2.19.0.md`; a `/scholar:hub skills` row in the Reference Card and the hub tutorial.
+
+### Changed
+
+- `scaffolded-math-tutorial` gains a 4th Terminology-Closure-Audit rule — **one symbol, one meaning; flag context-dependent forms** — guarding against attaching one definition of a symbol (e.g. a degrees-of-freedom that differs between a pooled-$t$ and a pooled-LRT reference) to a reference that needs a different form.
+
+### Fixed
+
+- `markdown-link-check`: ignore `imsglobal.org` and `opensource.org/licenses` — valid external links that intermittently return Status 0 / 403 to the checker and flaked the docs CI.
+
 ## [2.18.1] — 2026-06-12
 
 ### Security
@@ -1386,6 +1403,7 @@ text.replace(/\$([^$]+)\$/g, (match, content) => {
 
 ---
 
+[2.19.0]: https://github.com/Data-Wise/scholar/compare/v2.18.1...v2.19.0
 [2.18.1]: https://github.com/Data-Wise/scholar/compare/v2.18.0...v2.18.1
 [2.18.0]: https://github.com/Data-Wise/scholar/compare/v2.17.0...v2.18.0
 [2.17.0]: https://github.com/Data-Wise/scholar/compare/v2.16.0...v2.17.0

diff --git a/CLAUDE.md b/CLAUDE.md
@@ -54,15 +54,15 @@ This file contains project-specific instructions for Claude Code when working on
 ### Overview
 Scholar is a Claude Code plugin for academic workflows with research + teaching capabilities.
 
-### Current State (v2.18.1 released 2026-06-12)
-- **Research:** 14 commands, 17 skills
+### Current State (v2.19.0 released 2026-06-13)
+- **Research:** 14 commands, 18 skills (4 mathematical + 5 implementation + 3 writing + 5 research + 1 teaching)
 - **Teaching:** 18 commands (`/teaching:exam`, `/teaching:quiz`, `/teaching:slides`, `/teaching:assignment`, `/teaching:solution`, `/teaching:syllabus`, `/teaching:rubric`, `/teaching:feedback`, `/teaching:demo`, `/teaching:lecture`, `/teaching:validate`, `/teaching:validate-r`, `/teaching:diff`, `/teaching:sync`, `/teaching:migrate`, `/teaching:config`, `/teaching:preflight`, `/teaching:canvas`)
 - **Hub:** 1 command (`/scholar:hub` — command discovery and navigation, with flag discovery)
-- **Tests:** 3,405 (3,268 Jest + 137 node:test), 143 suites
+- **Tests:** 3,411 (3,268 Jest + 143 node:test), 144 suites
 - **Tutorials:** 38 (research-namespace coverage ~100%, teaching-namespace 100%; indexed at `docs/tutorials/index.md`)
 - **Docs:** MkDocs site on GitHub Pages — https://Data-Wise.github.io/scholar/
 - **Integration:** Works with flow-cli for workflow automation
-- **Release:** https://github.com/Data-Wise/scholar/releases/tag/v2.18.1 — maintenance: dep CVE patch (0 vulns), ESLint gate, `\printanswers` short-answer fix, timing de-flake
+- **Release:** https://github.com/Data-Wise/scholar/releases/tag/v2.19.0 — `scaffolded-math-tutorial` teaching skill (18th); skills surfaced in the hub (`/scholar:hub skills` + discovery `discoverSkills`); `skill_count` macro; df-conflation closure-audit guardrail; markdown-link-check flake fixes
 - **Active branch:** `dev` (clean)
 - **Next:** rich-exam v1 grading-line gap (`.STATUS` backlog #13); rich-exam v2 examark QTI (backlog #11)
 

diff --git a/README.md b/README.md
@@ -5,8 +5,8 @@
 A comprehensive Claude Code plugin for academic workflows combining research and teaching. Features unified Plugin + MCP architecture with 33 slash commands and research skills.
 
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
-[![Version](https://img.shields.io/badge/version-2.18.1-blue.svg)](https://github.com/Data-Wise/scholar/releases/tag/v2.18.1)
-[![Tests](https://img.shields.io/badge/tests-3,405%20passing-brightgreen.svg)](https://github.com/Data-Wise/scholar)
+[![Version](https://img.shields.io/badge/version-2.19.0-blue.svg)](https://github.com/Data-Wise/scholar/releases/tag/v2.19.0)
+[![Tests](https://img.shields.io/badge/tests-3,411%20passing-brightgreen.svg)](https://github.com/Data-Wise/scholar)
 [![Scholar CI](https://github.com/Data-Wise/scholar/actions/workflows/ci.yml/badge.svg?branch=main)](https://github.com/Data-Wise/scholar/actions/workflows/ci.yml?query=branch%3Amain)
 [![Docs Deploy](https://github.com/Data-Wise/scholar/actions/workflows/docs.yml/badge.svg?branch=main)](https://github.com/Data-Wise/scholar/actions/workflows/docs.yml?query=branch%3Amain)
 
@@ -113,7 +113,7 @@ This architecture eliminates IPC overhead by sharing core logic directly between
 - `src/teaching/config/` - Configuration management
 - `src/teaching/validators/` - Multi-layer validation (Schema + LaTeX + Completeness)
 - `src/teaching/ai/` - AI content generation with retry logic
-- `tests/teaching/` - 3,405 unit tests (100% passing)
+- `tests/teaching/` - 3,411 unit tests (100% passing)
 
 See [Phase 0 Architecture](docs/architecture/PHASE-0-FOUNDATION.md) for detailed documentation.
 
@@ -142,9 +142,9 @@ The Homebrew formula automatically:
 - Makes it available in Claude Code CLI and Claude Desktop
 - No additional configuration needed
 
-**Latest version:** v2.18.1 (released 2026-02-09)
+**Latest version:** v2.19.0 (released 2026-02-09)
 - 33 commands (18 teaching + 14 research + 1 hub)
-- 3,405 tests with 100% pass rate
+- 3,411 tests with 100% pass rate
 - Comprehensive documentation (95% coverage)
 
 ### Option 2: Manual Installation (Local Development)
@@ -198,7 +198,7 @@ cd ~/projects/dev-tools/scholar
 - 33 commands present (18 teaching + 14 research + 1 hub)
 - 17 skills present
 - No hardcoded paths
-- v2.18.1 verified
+- v2.19.0 verified
 ```
 
 ### Using in Claude Code CLI

diff --git a/docs/API-REFERENCE.md b/docs/API-REFERENCE.md
@@ -4,7 +4,7 @@ render_macros: false
 
 # Scholar Plugin - API Reference
 
-> **Version:** 2.18.1
+> **Version:** 2.19.0
 > **Last Updated:** 2026-03-04
 > **Audience:** Plugin developers and contributors
 >

diff --git a/docs/REFCARD.md b/docs/REFCARD.md
@@ -218,9 +218,11 @@
 | `/scholar:hub`   | Browse all commands    | `/scholar:hub`                    |
 | `/scholar:hub`   | Drill into category    | `/scholar:hub research`           |
 | `/scholar:hub`   | Command detail         | `/scholar:hub exam`               |
+| `/scholar:hub`   | List skills            | `/scholar:hub skills`             |
 | `/scholar:hub`   | Quick reference card   | `/scholar:hub quick`              |
 
 - **NEW in v2.7.0:** `/scholar:hub` for command discovery with smart context detection
+- **NEW:** `/scholar:hub skills` lists the 18 auto-activating skills grouped by category
 
 ---
 

diff --git a/docs/TEACHING-WORKFLOWS.md b/docs/TEACHING-WORKFLOWS.md
@@ -4,7 +4,7 @@ render_macros: false
 
 # Scholar Teaching Workflows
 
-> **Version:** 2.18.1
+> **Version:** 2.19.0
 > **Last Updated:** 2026-02-09
 > **Audience:** Educators using Scholar for course material generation
 >
@@ -958,5 +958,5 @@ Regenerate with new settings.
 ---
 
 **Last updated:** 2026-02-09
-**Version:** 2.18.1
+**Version:** 2.19.0
 **Contribute your workflows:** [GitHub Issues](https://github.com/Data-Wise/scholar/issues)
diff --git a/docs/WHATS-NEW-v2.18.0.md b/docs/WHATS-NEW-v2.18.0.md
@@ -46,5 +46,4 @@ The generated `.qmd` compiles with `quarto render exam.qmd --to pdf` using `lual
 ## Related
 
 - [Spec: Rich Exam Format](specs/SPEC-2026-05-12-rich-exam-format.md)
-- [ORCHESTRATE: exam-rich-latex](https://github.com/Data-Wise/scholar/blob/feature/exam-rich-latex/ORCHESTRATE-exam-rich-latex.md)
 - [CHANGELOG v2.18.0](https://github.com/Data-Wise/scholar/blob/main/CHANGELOG.md#2180--2026-05-12)
diff --git a/docs/WHATS-NEW-v2.19.0.md b/docs/WHATS-NEW-v2.19.0.md
@@ -0,0 +1,37 @@
+# What's New in v2.19.0
+
+**Status:** In development (targets v2.19.0)
+
+This release adds a new teaching skill and makes **all** skills discoverable from the hub.
+
+## scaffolded-math-tutorial skill
+
+A new auto-activating skill for turning a paper, proof, or formula into a **self-study course** a motivated reader (college-algebra floor) can learn from alone. It encodes a proven methodology:
+
+- **Four pillars** — zero undefined jargon + a running glossary; prerequisite primers taught before use; self-explanation + faded worked examples as the backbone (quizzing as a supplement); and a per-section teaching scaffold (objectives → worked instance → recap).
+- **The Terminology Closure Audit** — the enforceable version of "define every term": a compound-term rule, a no-forward-reference (closure) rule, and a mandated final audit pass.
+- **Dual HTML/PDF Quarto** output with folded-answer checkpoints.
+
+It brings the skill count to **18** and adds a new **Teaching** category to the [Skills Reference](skills.md). Background and the learning-science basis live in the skill's `reference/pedagogy.md`.
+
+## Skills in the hub
+
+The hub now surfaces **skills** alongside commands:
+
+```bash
+/scholar:hub          # overview now includes a SKILLS block (per-category counts)
+/scholar:hub skills   # all auto-activating skills, grouped by category, with inline help
+```
+
+Under the hood, the discovery engine gained `discoverSkills()` / `getSkillStats()` / `getSkillsByCategory()`, scanning `skills/**/SKILL.md`. The change is additive — the command-scan path and cache are untouched, so there is no behavior change for existing commands.
+
+Unlike commands, skills are **not invoked** — they activate automatically when your conversation matches their description.
+
+## Docs & tooling
+
+- New **Teaching** category in the [Skills Reference](skills.md); the [Reference Card](REFCARD.md) lists `/scholar:hub skills`; the [hub tutorial](tutorials/getting-started-with-hub.md) shows how to list skills.
+- A `skill_count` macro is now tracked in `mkdocs.yml` and synced by `scripts/version-sync.js`, so the skill total stays consistent across docs.
+
+## Notes
+
+- Skills are not command-count-gated; CI's "≥15 skills" floor still holds.
diff --git a/docs/help/FAQ-teaching.md b/docs/help/FAQ-teaching.md
@@ -2,7 +2,7 @@
 render_macros: false
 ---
 
-# Teaching FAQ - Scholar v2.18.1
+# Teaching FAQ - Scholar v2.19.0
 
 > **Quick answers to common questions about Scholar's teaching commands**
 
@@ -1768,4 +1768,4 @@ See [Migration Guide](../MIGRATION-v2.2.0.md).
 
 ---
 
-**Last Updated:** 2026-05-12 for Scholar v2.18.1
+**Last Updated:** 2026-05-12 for Scholar v2.19.0
diff --git a/docs/help/TROUBLESHOOTING-teaching.md b/docs/help/TROUBLESHOOTING-teaching.md
@@ -5,7 +5,7 @@ render_macros: false
 # Troubleshooting Guide - Scholar Teaching Commands
 
 > **Quick Fixes for Common Issues**
-> Version: 2.18.1 | Last Updated: 2026-05-12
+> Version: 2.19.0 | Last Updated: 2026-05-12
 
 ---
 
@@ -38,7 +38,7 @@ scholar --version
 **Expected output:**
 
 ```
-Scholar v2.18.1
+Scholar v2.19.0
 ```
 
 **If not found:**
@@ -574,7 +574,7 @@ claude plugin install scholar
 
 ```bash
 # Check version
-scholar --version    # Should be 2.18.1+
+scholar --version    # Should be 2.19.0+
 
 # Update if older
 brew upgrade scholar
@@ -2642,7 +2642,7 @@ head -40 exam.tex          # YAML preamble should match what's in the .qmd
 | `! Class exam Error` | `exam` class not installed | Install `texlive-latex-extra` (Linux) or `mactex-extra` (macOS) |
 | `! Undefined control sequence \printanswers` | Trying to render answer key without `exam` class | Same fix — install `texlive-latex-extra` |
 | `Quarto not found` | Quarto CLI missing | Install Quarto 1.4+ from <https://quarto.org/docs/get-started/> |
-| Bracket table shows `NaN` rows | Invalid df pair supplied (`df1: 0`, etc.) | Both `df1` and `df2` must be finite integers ≥ 1; Scholar v2.18.1+ throws on invalid input |
+| Bracket table shows `NaN` rows | Invalid df pair supplied (`df1: 0`, etc.) | Both `df1` and `df2` must be finite integers ≥ 1; Scholar v2.19.0+ throws on invalid input |
 
 **Step 1: Verify the YAML preamble**
 
@@ -3379,7 +3379,7 @@ If custom prompts are complex, use refine:
 
 **Solution:**
 
-1. File was generated before provenance tracking (requires Scholar v2.18.1+)
+1. File was generated before provenance tracking (requires Scholar v2.19.0+)
 2. File was not generated by Scholar (e.g., manually created)
 3. Frontmatter was modified and metadata comments were removed
 4. Re-generate the file with current Scholar version to add provenance
@@ -4315,4 +4315,4 @@ Visit: https://console.anthropic.com/account/billing/usage
 
 **Questions? Spotted an issue?** Open a GitHub issue: https://github.com/Data-Wise/scholar/issues
 
-**Last Updated:** 2026-05-12 for Scholar v2.18.1
+**Last Updated:** 2026-05-12 for Scholar v2.19.0
diff --git a/docs/skills.md b/docs/skills.md
@@ -1,13 +1,14 @@
 # Scholar Skills Reference
 
-> **Version {{ scholar.version }}** - 17 A-grade auto-activating skills for statistical research and methods development
+> **Version {{ scholar.version }}** - {{ scholar.skill_count }} A-grade auto-activating skills for statistical research, methods development, and teaching
 
 ## Skill Categories
 
 - **[Mathematical](#mathematical-skills)** (4 skills)
 - **[Implementation](#implementation-skills)** (5 skills)
 - **[Writing](#writing-skills)** (3 skills)
 - **[Research](#research-skills)** (5 skills)
+- **[Teaching](#teaching-skills)** (1 skill)
 
 ## Mathematical Skills
 
@@ -238,6 +239,22 @@ Sensitivity analysis design.
 
 **Activates on:** "sensitivity", "robustness", "assumptions"
 
+## Teaching Skills
+
+### scaffolded-math-tutorial
+
+Self-study math/stats tutorials that teach formulas and derivations from the ground up — a
+motivated reader (college-algebra floor) can learn from the document alone.
+
+**Expertise:**
+
+- College-algebra-floor scaffolding (prerequisite primers taught before use)
+- The Terminology Closure Audit (zero undefined jargon; no forward references)
+- Self-explanation + faded worked examples as the backbone (quizzing as supplement)
+- Folded-answer checkpoints; dual HTML (folded) / PDF (open) Quarto rendering
+
+**Activates on:** "tutorial", "teach me the math", "explain this derivation", "scaffold a proof"
+
 ## Usage Examples
 
 ### Automatic Activation

diff --git a/docs/tutorials/getting-started-with-hub.md b/docs/tutorials/getting-started-with-hub.md
@@ -237,6 +237,14 @@ Tip: You're in a teaching project. Try /teaching:exam or /scholar:hub for all co
 
 Look for `[AI]` markers in the overview, or drill into a command to check for `--instructions / -i` in its Options section.
 
+### "What skills do I have?"
+
+```bash
+/scholar:hub skills    # All auto-activating skills, grouped by category
+```
+
+Skills aren't invoked like commands -- they activate automatically when your conversation matches their description. See the [Skills Reference](../skills.md) for the full list.
+
 ### "I'm new -- where do I start?"
 
 ```bash

diff --git a/docs/workflows/research/latex-integration.md b/docs/workflows/research/latex-integration.md
@@ -1031,7 +1031,7 @@ Week 4: Advanced automation (Makefiles, CI/CD)
 
 ---
 
-**Document Version:** v2.18.1
+**Document Version:** v2.19.0
 **Last Updated:** 2026-02-01
 **Word Count:** ~7,500
 **Example Scripts:** 15+
diff --git a/docs/workflows/research/literature-review.md b/docs/workflows/research/literature-review.md
@@ -1249,7 +1249,7 @@ After completing literature review:
 
 ---
 
-**Document Version:** v2.18.1
+**Document Version:** v2.19.0
 **Last Updated:** 2026-02-01
 **Word Count:** ~9,500
 **Time to Complete:** 2-4 weeks for systematic review
diff --git a/docs/workflows/research/manuscript-writing.md b/docs/workflows/research/manuscript-writing.md
@@ -1813,7 +1813,7 @@ pandoc manuscript.md --bibliography references.bib -o manuscript.pdf
 
 ---
 
-**Document Version:** 2.18.1
+**Document Version:** 2.19.0
 **Last Updated:** 2026-02-04
 **Author:** Data-Wise Team
 **Status:** Complete

diff --git a/mkdocs.yml b/mkdocs.yml
@@ -44,17 +44,18 @@ extra_css:
 
 extra:
   scholar:
-    version: "2.18.1"
-    prev_version: "2.18.0"
-    release_date: "2026-06-12"
-    test_count: "3,405"
+    version: "2.19.0"
+    prev_version: "2.18.1"
+    release_date: "2026-06-13"
+    test_count: "3,411"
     jest_count: "3,268"
-    node_test_count: "137"
+    node_test_count: "143"
     command_count: 33
     teaching_commands: 18
     research_commands: 14
+    skill_count: 18
     hub_commands: 1
-    suite_count: 143
+    suite_count: 144
 
 markdown_extensions:
   - pymdownx.highlight:
@@ -238,6 +239,7 @@ nav:
           - Phase 0 Foundation: architecture/PHASE-0-FOUNDATION.md
 
   - Release Notes:
+      - What's New in v2.19.0: WHATS-NEW-v2.19.0.md
       - What's New in v2.18.0: WHATS-NEW-v2.18.0.md
       - What's New in v2.17.0: WHATS-NEW-v2.17.0.md
       - What's New in v2.16.0: WHATS-NEW-v2.16.0.md