kenmark-skills 1.0.0

package/skills/user-skills/issues-setup/SKILL.md

@@ -0,0 +1,176 @@
+---
+name: issues-setup
+preamble-tier: 2
+version: 1.0.0
+description: |
+  Set up the brain/issues/ directory structure: create brain/issues/,
+  brain/issues/completed/, and brain/issues/INDEX.md from scratch. Use when
+  asked to "setup issues", "init issues", or "bootstrap issues directory".
+allowed-tools:
+  - Bash
+  - Read
+  - Write
+  - Edit
+  - Grep
+  - Glob
+  - AskUserQuestion
+triggers:
+  - setup issues
+  - init issues
+  - bootstrap issues directory
+  - initialize issues
+---
+
+# Issues Setup — Bootstrap brain/issues/ Directory
+
+## Purpose
+
+Create the full `brain/issues/` directory structure from scratch:
+- `brain/issues/` — active issues
+- `brain/issues/completed/` — archived resolved issues
+- `brain/issues/INDEX.md` — master index with all issues grouped by priority
+
+---
+
+## Step 1 — Find or create the brain directory
+
+```bash
+REPO_ROOT="$(git rev-parse --show-toplevel 2>/dev/null)"
+ISSUES_DIR="$REPO_ROOT/brain/issues"
+COMPLETED_DIR="$ISSUES_DIR/completed"
+echo "REPO_ROOT=$REPO_ROOT"
+echo "ISSUES_DIR=$ISSUES_DIR"
+```
+
+If `brain/` does not exist, create it:
+
+```bash
+mkdir -p "$ISSUES_DIR" "$COMPLETED_DIR"
+echo "Created: $ISSUES_DIR and $COMPLETED_DIR"
+```
+
+---
+
+## Step 2 — Create the INDEX.md
+
+Write `brain/issues/INDEX.md` with this exact structure:
+
+```markdown
+# Issues Index
+
+## Overview
+
+| Category | Count |
+|----------|-------|
+| Active issues | 0 |
+| Completed | 0 |
+| **Total** | **0** |
+
+## Completed Issues
+
+| ID | Title | Completed |
+|----|-------|-----------|
+| _none yet_ | | |
+
+## Active Issues by Priority
+
+### P0 — Critical
+
+| ID | Title |
+|----|-------|
+| _none_ | |
+
+### P1 — High
+
+| ID | Title |
+|----|-------|
+| _none_ | |
+
+### P2 — Medium
+
+| ID | Title |
+|----|-------|
+| _none_ | |
+
+## Issue Structure
+
+Each issue file contains:
+
+\`\`\`yaml
+---
+id: XXX
+title: ...
+severity: P0|P1|P2
+area: api|database|security|ui|worker|testing|maintainability|connectors|oauth|dsl
+source: how-the-issue-was-found
+status: open|completed
+created: YYYY-MM-DD
+files:
+  - relevant-files
+related:
+  - related-issue-ids
+---
+
+## Summary
+
+## Evidence
+
+## Suggested fix
+
+## Acceptance criteria
+\`\`\`
+
+## Areas
+
+| Area | Description |
+|------|-------------|
+| api | REST API routes and handlers |
+| database | Prisma schema, migrations |
+| security | Auth, HMAC, rate limiting, input validation |
+| ui | React components, pages, navigation |
+| worker | Background job processing |
+| testing | Test coverage gaps |
+| maintainability | Code quality, dead code |
+| connectors | Connector implementations |
+| oauth | OAuth flows, Google integration |
+| dsl | Workflow DSL schema and types |
+
+## Workstreams
+
+| Workstream | Issues |
+|-----------|--------|
+| Security |016, 017, 018, 019, 023, 024, 026 |
+| Worker | 012, 022, 025, 027 |
+| Database migrations | 013, 014, 021 |
+| DSL vocabulary | 015 |
+| UI pages rename | 001, 003, 007, 008, 038, 039, 040, 041, 047, 048, 049, 050, 051, 052, 053, 054, 055, 056, 057, 058 |
+| UI component naming | 009, 010, 011 |
+| Test gaps | 028, 030, 031, 032, 033, 034, 035, 036, 037 |
+| Accessibility | 042, 043, 044, 045, 046, 052, 054, 055 |
+```
+
+---
+
+## Step 3 — Verify structure
+
+```bash
+echo "=== Directory structure ==="
+ls -la "$ISSUES_DIR/"
+echo ""
+echo "=== Completed dir ==="
+ls -la "$COMPLETED_DIR/"
+echo ""
+echo "=== INDEX.md exists ==="
+[ -f "$ISSUES_DIR/INDEX.md" ] && echo "YES" || echo "MISSING"
+```
+
+---
+
+## Step 4 — Confirm to user
+
+Report what was created:
+- `brain/issues/` directory
+- `brain/issues/completed/` directory
+- `brain/issues/INDEX.md` with full template
+
+Then suggest next steps: use `/issues-scan` to populate with issues discovered from the codebase, or use `/issues-list` to see the empty state.

package/skills/user-skills/skill-router/SKILL.md

@@ -0,0 +1,147 @@
+---
+name: skill-router
+description: Routes tasks to the best installed user skill by analyzing intent, keywords, and category. Use at the start of any task when unsure which skill applies, when the user invokes /skill-router, or when multiple domains overlap (SEO vs design vs backend vs testing).
+---
+
+# Skill Router
+
+Automatically pick and load the best **user skill** for the current task.
+
+## Canonical skill store
+
+All user skills live in `~/.agents/skills/`.
+
+This router uses a **local generated registry** at:
+`skills/user-skills/skill-router/registry.json`
+
+Do not commit this file. It is generated per-user on first run and refreshed on every call.
+
+**Do not modify project skills** (`<repo>/.agents/skills/`, `<repo>/.cursor/skills/`). Those are repo-scoped and out of scope for this router.
+
+## Runtime bootstrap (always run first)
+
+Before scoring skills, regenerate the local registry file:
+
+```bash
+python3 - <<'PY'
+import json
+import re
+from datetime import datetime, timezone
+from pathlib import Path
+
+repo_root = Path.cwd()
+out_path = repo_root / "skills/user-skills/skill-router/registry.json"
+user_root = Path.home() / ".agents/skills"
+
+def infer_category(name: str, description: str) -> str:
+    text = f"{name} {description}".lower()
+    if "seo" in text:
+        return "seo"
+    if any(k in text for k in ["design", "ui", "ux", "frontend", "visual"]):
+        return "design"
+    if any(k in text for k in ["test", "qa", "verification", "eval", "tdd"]):
+        return "testing"
+    if any(k in text for k in ["workflow", "router", "learning", "agent"]):
+        return "workflow"
+    if any(k in text for k in ["api", "backend", "django", "python", "database", "server"]):
+        return "backend"
+    return "general"
+
+def extract_description(skill_md: str) -> str:
+    m = re.search(r"(?ms)^description:\s*(.+?)(?:\n\w[\w-]*:|\n---|\Z)", skill_md)
+    if not m:
+        return ""
+    return re.sub(r"\s+", " ", m.group(1)).strip().strip('"').strip("'")
+
+def extract_keywords(name: str, description: str) -> list[str]:
+    tokens = re.findall(r"[a-zA-Z0-9][a-zA-Z0-9\-]{2,}", f"{name} {description}".lower())
+    seen = set()
+    out = []
+    for t in tokens:
+        if t not in seen:
+            seen.add(t)
+            out.append(t)
+    return out[:80]
+
+skills = []
+if user_root.exists():
+    for skill_dir in sorted([p for p in user_root.iterdir() if p.is_dir()]):
+        skill_md = skill_dir / "SKILL.md"
+        if not skill_md.exists():
+            continue
+        content = skill_md.read_text(encoding="utf-8", errors="ignore")
+        desc = extract_description(content)
+        name = skill_dir.name
+        skills.append({
+            "name": name,
+            "path": str(skill_dir),
+            "description": desc,
+            "category": infer_category(name, desc),
+            "keywords": extract_keywords(name, desc),
+        })
+
+payload = {
+    "version": 1,
+    "generatedAt": datetime.now(timezone.utc).isoformat(),
+    "canonicalRoot": str(user_root),
+    "skillCount": len(skills),
+    "skills": skills,
+}
+
+out_path.parent.mkdir(parents=True, exist_ok=True)
+out_path.write_text(json.dumps(payload, indent=2, ensure_ascii=False) + "\n", encoding="utf-8")
+print(f"Generated {out_path} with {len(skills)} skills")
+PY
+```
+
+## When to run
+
+Run this skill **before** other skills when:
+
+- The user's request spans multiple domains
+- No skill was auto-selected but specialized guidance would help
+- The user says `/skill-router`, "pick the right skill", or "which skill should we use?"
+- You are about to guess workflow steps that an existing skill already encodes
+
+## Routing algorithm
+
+1. **Parse intent** from the user message: domain, deliverable, stack, and constraints.
+2. **Load registry**: read [registry.json](registry.json) after the runtime bootstrap above completes.
+3. **Score candidates** (top 3):
+   - +3 exact skill name or alias in the request; +2 keyword overlap with task
+   - +2 category match (seo, design, backend, testing, workflow)
+   - +1 trigger phrase from skill description appears in the task
+   - −2 category mismatch (e.g. SEO skill for a pure DB migration)
+4. **Tie-break**: prefer narrower skills over umbrella skills (`seo-page` over `seo`; `django-tdd` over `coding-standards`).
+5. **Load winner**: read the selected skill's `SKILL.md` and follow it for the rest of the task.
+6. **Multi-skill tasks**: if two skills score within 1 point and serve different phases (e.g. `ce-plan` + `tdd-workflow`), load primary first, note secondary for the next phase.
+7. **No match (score < 2)**: use `find-skills` to search for installable skills, then proceed with general capabilities.
+
+## Category quick map
+
+| If the task is about… | Prefer skills in category… | Examples |
+| --- | --- | --- |
+| Search, rankings, metadata, structured data | `seo` | `seo-audit`, `seo-technical`, `seo-schema` |
+| UI polish, layout, visual design, MUI | `design` | `impeccable`, `design-taste-frontend`, `high-end-visual-design` |
+| APIs, services, frameworks, languages | `backend` | `backend-patterns`, `django-patterns`, `python-patterns` |
+| Tests, QA, verification, evals | `testing` | `tdd-workflow`, `verification-loop`, `eval-harness` |
+| Agent workflow, discovery, learning | `workflow` | `find-skills`, `continuous-learning`, `skill-router` |
+
+## Output format
+
+After routing, tell the user briefly:
+
+```markdown
+**Routed skill:** `<name>` (<category>)
+**Why:** <one sentence>
+**Next:** <first concrete step from that skill>
+```
+
+Then execute using the routed skill — do not stop at the recommendation.
+
+## Maintenance
+
+- Add or edit skills only under `~/.agents/skills/<skill-name>/`
+- Do not commit `skills/user-skills/skill-router/registry.json` (generated per-user file)
+- Runtime bootstrap refreshes `registry.json` on each invocation
+- Install new skills globally with: `npx skills add <owner/repo@skill> -g -y --agent claude-code cursor codex`