wicked-garden 12.27.0

package/skills/classify/SKILL.md

@@ -0,0 +1,157 @@
+---
+name: wicked-garden-classify
+context: fork
+description: |
+  v11 LLM-based work-shape classifier. Replaces the regex archetype detector
+  with the model's own reasoning. Reads the user's prompt, picks the right
+  archetype(s) from the catalog, identifies signals (blast_radius, novelty,
+  reversibility, etc.), and persists to SessionState so subsequent turns
+  steer correctly.
+
+  Use when: the prompt_submit hook emitted a `<wg classify-due />` directive,
+  OR explicitly invoked at session start, OR when re-classifying after the
+  user changes scope mid-session.
+allowed-tools: ["*"]
+---
+
+# /wicked-garden:classify
+
+You are classifying a prompt into a v11 work-shape archetype. Your output
+drives downstream archetype routing for the rest of the session (until
+the user changes scope or invokes this skill again).
+
+## Why this exists
+
+The v11 hook auto-classifier is a regex + boolean-signal heuristic. It
+works for prompts whose vocabulary matches the catalog phrase lists, but
+it misses paraphrases and underuses signals. **You are a better classifier
+than regex.** This skill is the path to use the model's full reasoning
+on the classification step, then persist the result so the rest of the
+session benefits without re-running classification on every turn.
+
+## What the catalog declares
+
+`.claude-plugin/archetypes.json` defines the work-shape archetypes. Read it
+once at the start of this skill. Summary table:
+
+| Archetype | Phases                                        | Use when                                          |
+|-----------|-----------------------------------------------|---------------------------------------------------|
+| triage    | classify                                      | prompt is genuinely ambiguous; ask for clarification |
+| explore   | frame → diverge → converge                    | open problem space, multiple paths, brainstorm    |
+| specify   | elicit → structure → validate                 | requirements / acceptance criteria need writing   |
+| decide    | brief → options → score → record              | 2+ viable options, need an ADR                    |
+| ship      | canary → ramp → full → soak                   | already-built change being rolled out             |
+| review    | scope → assess → findings → remediate-or-accept | independent assessment of an artifact            |
+| incident  | triage → investigate → mitigate → resolve → followup | live production failure                      |
+| build     | plan → implement → test → review              | implement a feature or fix (most common)          |
+| migrate   | plan → expand → backfill → cutover → contract | in-place shape change with rollback proof         |
+| modernize | discover → extract → blueprint → transform → parity → cutover | port a legacy codebase to a new stack (NOT in-place) |
+
+## Procedure
+
+### 1. Read the prompt
+
+What is the user actually asking for? Restate in one sentence in your own
+words. If you can't, the prompt is genuinely ambiguous → triage.
+
+### 2. Pick archetype(s)
+
+Archetypes are NOT mutually exclusive. Pick a SET. Common combinations:
+
+- "implement schema change with backfill" → `build + migrate`
+- "review the auth PR before deploy" → `review + ship`
+- "should we use redis or memcached for sessions?" → `decide` (and
+  possibly `build` if they want you to also implement it)
+
+Score each match between 0.0 and 1.0. Use these calibration anchors:
+
+- **0.9+**: keyword + signal both clear, no ambiguity (e.g. "checkout is
+  down 5xx spiking" → incident 1.0).
+- **0.7–0.9**: clear shape with one or two minor uncertainties.
+- **0.5–0.7**: archetype matches but the prompt is partial — agent should
+  read the playbook and gracefully ask if anything was missed.
+- **<0.5**: too weak; don't return this archetype.
+
+If nothing scores ≥ 0.5, return `triage` only — that's the signal to ask
+for clarification before doing work.
+
+### 3. Identify signals
+
+Boolean flags that downstream archetypes use to scale rigor. Mark TRUE
+only when the prompt clearly implies it:
+
+- `blast_radius_high` — change affects production traffic / many users / many systems.
+- `novelty_high` — pattern not yet in this codebase.
+- `state_complexity_high` — touches data shape, migrations, persistent state.
+- `reversibility_low` — undoing is expensive (data migrations, destructive ops).
+- `reversibility_medium_or_low` — undoing is non-trivial (config changes, breaking APIs).
+- `production_impact` — production users / systems affected right now.
+- `compliance_scope` — GDPR / SOC2 / HIPAA / PCI surface.
+- `ambiguity_high` — multiple plausible reads.
+- `spec_ambiguity_high` — success criteria are fuzzy.
+- `scope_unclear` — boundary of work is undefined.
+- `multiple_viable_options` — 2+ paths with no obvious winner.
+- `post_build` — change is already implemented; this is about deployment.
+- `code_change` — implementation work involved.
+- `independent_assessment_needed` — someone else's work needs review.
+
+Default any flag you didn't explicitly mark to FALSE. Do not over-tag.
+
+### 4. Pick intent
+
+Intent is coarser than archetype — used by the hook to gate directive
+emission. One of:
+
+- `simple-edit` — typo, comment, formatting, single-line fix. Hook stays silent.
+- `feature` — most non-trivial work (default for build/migrate/ship).
+- `rigor` — high stakes (compliance, security, blast_radius_high).
+- `research` — exploratory (explore, decide).
+
+### 5. Persist
+
+Emit a JSON object with the four keys above and pipe to the persist
+script. Use this exact shape — extras get dropped:
+
+```bash
+echo '{
+  "intent": "feature",
+  "archetypes": [
+    {"name": "build", "score": 0.85, "evidence": ["implement keyword + code_change signal"]},
+    {"name": "migrate", "score": 0.65, "evidence": ["schema change + state_complexity signal"]}
+  ],
+  "signals": {
+    "code_change": true,
+    "state_complexity_high": true,
+    "reversibility_low": true
+  }
+}' | sh "${CLAUDE_PLUGIN_ROOT}/scripts/_python.sh" \
+  "${CLAUDE_PLUGIN_ROOT}/scripts/classify/persist.py"
+```
+
+The script normalises and writes to SessionState. Confirm the response
+shows `"ok": true`, then return control.
+
+### 6. Hand off
+
+After persisting, look at the top archetype's playbook
+(`skills/archetype/refs/{name}.md`) and start running it. Do not re-run
+classification mid-session unless the user explicitly changes scope.
+
+## When to skip this skill
+
+- The prompt is a continuation token ("yes", "do it", "lgtm"). The hook
+  already short-circuits these.
+- The user typed `/wicked-garden:archetype:<name>` directly — they
+  already classified.
+- SessionState already has `classified_at` set for this session and the
+  prompt fits the existing classification. Re-classifying on every turn
+  is exactly the cost we're trying to avoid.
+
+## Failure modes
+
+- Persist script fails: return the JSON to the user as text and ask
+  them how to proceed. Do not invent a workflow.
+- Catalog file unreadable: return triage with an explanation. The
+  archetype machinery will surface the read error separately.
+- Model can't decide between 3+ archetypes: return triage with a note.
+  triage's job is to ask. Don't pick at random.

package/skills/data/analysis/SKILL.md

@@ -0,0 +1,125 @@
+---
+name: analysis
+description: |
+  Use when exploring a dataset for patterns, trends, and business insights — EDA, segmentation,
+  anomaly detection, and visualization guidance. Generates the Observation → Insight → Action chain.
+  NOT for schema validation or data quality scoring (use the data/data skill) or SQL queries (use data:analyze).
+phase_relevance: ["design", "build"]
+archetype_relevance: ["*"]
+---
+
+# Data Analysis Skill
+
+Explore datasets, identify patterns, and generate actionable insights.
+
+## Quick Start
+
+```bash
+/wicked-garden:data:analyze explore data/sales.csv
+```
+
+This will:
+1. Profile the data
+2. Generate descriptive statistics
+3. Identify patterns and trends
+4. Suggest visualizations
+5. Provide initial insights
+
+After exploration, ask specific questions:
+- "What's the trend in sales over time?"
+- "Which customer segments are most valuable?"
+- "Are there any anomalies in the data?"
+
+## Analysis Workflow
+
+### 1. Profile & Understand
+
+```bash
+/wicked-garden:data:analyze sales.csv
+```
+
+**Key questions**:
+- What's the grain? (one row per what?)
+- What's the date range?
+- What are the key metrics?
+- Any obvious data quality issues?
+
+### 2. Explore Patterns
+
+**Basic explorations**:
+- Distributions (histograms, percentiles)
+- Categorical breakdowns (frequency tables)
+- Time trends (daily/monthly aggregations)
+- Correlations (relationship between metrics)
+
+### 3. Segment Analysis
+
+**Common segmentations**:
+- **RFM**: Recency, Frequency, Monetary value
+- **Cohorts**: Group by signup/first purchase month
+- **Geographic**: By region, country, city
+- **Behavioral**: By usage patterns
+
+### 4. Anomaly Detection
+
+**Detection methods**:
+- Statistical outliers (>3 standard deviations)
+- Unusual patterns (volume spikes/drops)
+- Temporal anomalies (day-of-week deviations)
+
+## Insight Generation
+
+### Pattern: Observation → Insight → Action
+
+**Bad** (just observations):
+- Average order value is $45
+- 60% of customers are in the US
+
+**Good** (insights with actions):
+```
+### Insight: Weekend Shopping Behavior
+**Observation**: Sales peak on Friday (+40% vs weekday avg)
+**Insight**: Customers shop for weekend needs on Friday.
+**Action**: Launch "Friday Flash Sale" campaign (+15% expected)
+**Confidence**: HIGH (consistent over 6 months)
+```
+
+## Visualization Guidance
+
+| Data Type | Comparison | Best Chart |
+|-----------|------------|------------|
+| Time series | Trend over time | Line chart |
+| Categorical | Compare values | Bar chart |
+| Distribution | Show spread | Histogram, Box plot |
+| Relationship | Correlation | Scatter plot |
+| Composition | Part of whole | Stacked bar, Pie |
+| Geographic | Location data | Map (choropleth) |
+
+## Integration
+
+**wicked-garden:data:analyze** - Primary tool for data queries via DuckDB:
+```bash
+/wicked-garden:data:analyze data.csv
+```
+
+**Native tasks** - Document insights via TaskCreate with `metadata.event_type="task"`
+**wicked-brain:memory** - Store analysis patterns for reuse
+
+## Best Practices
+
+- **Start simple**: Begin with basic aggregations before complex models
+- **Validate assumptions**: Check data quality, verify calculations
+- **Tell a story**: Lead with the "so what?" and make it actionable
+- **Be honest about uncertainty**: Note limitations, quantify confidence
+
+## Common Pitfalls
+
+- **Correlation ≠ Causation**: Ice cream and drownings both increase in summer
+- **Simpson's Paradox**: Aggregate trend may reverse when segmented
+- **Survivorship Bias**: Include failed cases for balanced view
+- **Cherry-picking**: Report all patterns, not just favorable ones
+
+## Reference
+
+For detailed content:
+- [Output Templates](refs/templates.md) - Report and insight templates

package/skills/data/analysis/refs/templates.md

@@ -0,0 +1,170 @@
+---
+phase_relevance: ["design", "build"]
+archetype_relevance: ["*"]
+---
+# Analysis Output Templates
+
+Standard templates for data analysis reports.
+
+## Exploratory Analysis Report
+
+```markdown
+## Exploratory Analysis: {dataset}
+
+**Date**: {date}
+**Analyst**: {name}
+**Rows**: {count}
+
+### Quick Stats
+- **Date Range**: {start} to {end}
+- **Key Metrics**: {summary}
+- **Data Quality**: {score}/100
+
+### Data Profile
+| Column | Type | Nulls | Unique | Min | Max |
+|--------|------|-------|--------|-----|-----|
+| {col}  | {type} | {pct}% | {n} | {min} | {max} |
+
+### Distributions
+{Summary of key metric distributions}
+
+### Key Findings
+1. {Finding with supporting data}
+2. {Finding with supporting data}
+3. {Finding with supporting data}
+
+### Data Quality Issues
+- {Issue and impact}
+
+### Questions Raised
+- {Question needing further investigation}
+
+### Next Steps
+- {Recommended follow-up analysis}
+```
+
+## Insight Report
+
+```markdown
+## Data Insights: {topic}
+
+**Context**: {why this analysis}
+**Data**: {source and timeframe}
+**Confidence**: {HIGH|MEDIUM|LOW}
+
+### Executive Summary
+{Top 3 takeaways in bullet points}
+
+---
+
+### Insight 1: {Title}
+
+**Observation**: {What the data shows - facts only}
+
+**Insight**: {What it means - interpretation}
+
+**Action**:
+- Specific action 1
+- Expected impact: {quantified if possible}
+
+**Confidence**: {HIGH|MEDIUM|LOW}
+
+---
+
+### Insight 2: {Title}
+{Same structure as above}
+
+---
+
+### Supporting Analysis
+{Methodology and data quality notes}
+
+### Assumptions & Limitations
+- {Key assumption}
+- {Data limitation}
+
+### Next Steps
+1. {Immediate action}
+2. {Follow-up analysis}
+```
+
+## Segment Analysis Report
+
+```markdown
+## Segment Analysis: {dimension}
+
+**Segments**: {n segments}
+**Method**: {RFM, cohort, behavioral, etc.}
+
+### Segment Overview
+| Segment | Size | % Total | Avg Value | Key Trait |
+|---------|------|---------|-----------|-----------|
+| {name}  | {n}  | {pct}%  | ${avg}    | {trait}   |
+
+### Segment Profiles
+
+#### Segment 1: {Name}
+- **Size**: {n} ({pct}%)
+- **Characteristics**: {description}
+- **Value**: ${total} ({pct}% of revenue)
+- **Recommended Action**: {action}
+
+### Cross-Segment Insights
+- {Pattern across segments}
+
+### Recommended Actions
+1. **{Segment}**: {Action}
+```
+
+## A/B Test Results
+
+```markdown
+## A/B Test Results: {experiment name}
+
+**Test Period**: {start} to {end}
+**Sample Size**: Control: {n}, Treatment: {n}
+
+### Summary
+| Metric | Control | Treatment | Lift | Confidence |
+|--------|---------|-----------|------|------------|
+| {KPI}  | {val}   | {val}     | +{x}%| {pct}%     |
+
+### Statistical Details
+- **Test Type**: {t-test, chi-square, etc.}
+- **p-value**: {p}
+- **Confidence Interval**: [{low}, {high}]
+
+### Recommendation
+{SHIP / DO NOT SHIP / NEEDS MORE DATA}
+
+### Caveats
+- {Limitation or consideration}
+```
+
+## Anomaly Report
+
+```markdown
+## Anomaly Detection: {metric}
+
+**Period**: {date range}
+**Method**: {statistical, isolation forest, etc.}
+
+### Anomalies Detected
+| Date | Value | Expected | Deviation | Severity |
+|------|-------|----------|-----------|----------|
+| {date} | {val} | {expected} | {+/-x}% | {HIGH/MED/LOW} |
+
+### Investigation
+
+#### Anomaly 1: {Date}
+- **Observed**: {value}
+- **Expected**: {value} (based on {method})
+- **Possible Causes**:
+  1. {Hypothesis}
+  2. {Hypothesis}
+- **Impact**: {business impact}
+- **Action**: {recommendation}
+
+### Systemic Issues
+- {Pattern in anomalies}
+```

package/skills/data/data/SKILL.md

@@ -0,0 +1,118 @@
+---
+name: data
+description: |
+  Use when profiling a dataset's structure, validating it against a schema, or generating a data quality
+  report (completeness, uniqueness, validity constraints). Runs the data_profiler.py and schema_validator.py scripts.
+  NOT for exploratory pattern analysis (use data/analysis) or SQL queries (use data:analyze).
+phase_relevance: ["design", "build"]
+archetype_relevance: ["*"]
+---
+
+# Data Engineering Skill
+
+Core data engineering operations for profiling, validation, and quality assessment.
+
+## Quick Start
+
+### Profile a Dataset
+
+```bash
+/wicked-garden:data:data profile path/to/data.csv
+```
+
+This will:
+1. Detect file format
+2. Sample rows (head/random/tail)
+3. Infer schema and types
+4. Calculate quality metrics
+5. Generate recommendations
+
+### Validate Schema
+
+```bash
+/wicked-garden:data:data validate --schema schema.json --data data.csv
+```
+
+Checks: Column presence, type conformance, constraint validation, nullability rules.
+
+### Assess Quality
+
+```bash
+/wicked-garden:data:data quality data.csv
+```
+
+Reports on: Completeness (null rates), Uniqueness (duplicates), Validity (constraints), Consistency (cross-field checks).
+
+## Commands
+
+| Command | Purpose |
+|---------|---------|
+| `/wicked-garden:data:data profile <path>` | Profile dataset structure and quality |
+| `/wicked-garden:data:data validate` | Validate data against schema |
+| `/wicked-garden:data:data quality <path>` | Generate quality report |
+
+## Dataset Profiling
+
+Uses `data_profiler.py` script:
+
+```bash
+sh "${CLAUDE_PLUGIN_ROOT}/scripts/_python.sh" "${CLAUDE_PLUGIN_ROOT}/scripts/data/data_profiler.py" \
+  --input data.csv --output profile.json
+```
+
+**Output includes**:
+- Row count and column count
+- Column types (inferred)
+- Null rates per column
+- Cardinality (distinct values)
+- Sample values
+- Statistical summaries for numeric columns
+
+## Schema Validation
+
+Uses `schema_validator.py` script. Define expected columns with:
+- **name**: Column name
+- **type**: integer, string, decimal, datetime, date
+- **nullable**: true/false
+- **constraints**: unique, min, max, pattern, enum
+
+See [examples](refs/examples.md) for schema format.
+
+## Quality Dimensions
+
+| Dimension | Metric | Threshold |
+|-----------|--------|-----------|
+| Completeness | Null rate | <5% |
+| Uniqueness | Duplicate rate | <1% |
+| Validity | Type conformance | 100% |
+| Consistency | Cross-field rules | 100% |
+
+## Integration
+
+| Plugin | Enhancement |
+|--------|-------------|
+| wicked-garden:data:analyze | Use for SQL-based profiling of large files via DuckDB |
+| Native tasks | Document quality issues via TaskCreate with `metadata.event_type="task"` |
+| wicked-brain:memory | Store quality patterns across sessions |
+
+## Large Files
+
+For files >1GB, use wicked-garden:data:analyze for efficient SQL-based profiling:
+
+```bash
+/wicked-garden:data:analyze large_file.csv
+```
+
+## Output Standards
+
+All reports include:
+- **Summary**: High-level findings
+- **Metrics**: Quantitative measurements
+- **Issues**: Prioritized problems
+- **Recommendations**: Actionable next steps
+- **Confidence**: Assessment reliability
+
+## Reference
+
+For detailed examples and patterns:
+- [Examples](refs/examples.md) - Profile output, schema format, quality report