specweave 1.0.550 → 1.0.552

package/plugins/specweave/hooks/llm-judge-validator.sh

@@ -1,219 +0,0 @@
-#!/bin/bash
-# llm-judge-validator.sh - LLM Judge Quality Assessment
-#
-# Uses Claude to assess work quality before allowing auto session completion
-# Returns exit code 0 if quality approved, 1 if rejected
-#
-# Usage: llm-judge-validator.sh <INCREMENT_ID> <TRANSCRIPT_PATH>
-
-set -e
-
-INCREMENT_ID="$1"
-TRANSCRIPT_PATH="$2"
-
-if [ -z "$INCREMENT_ID" ] || [ -z "$TRANSCRIPT_PATH" ]; then
-    echo "Usage: llm-judge-validator.sh <INCREMENT_ID> <TRANSCRIPT_PATH>" >&2
-    exit 1
-fi
-
-PROJECT_ROOT="${PROJECT_ROOT:-$(pwd)}"
-INCREMENT_DIR="$PROJECT_ROOT/.specweave/increments/$INCREMENT_ID"
-
-if [ ! -d "$INCREMENT_DIR" ]; then
-    echo "❌ Increment directory not found: $INCREMENT_DIR" >&2
-    exit 1
-fi
-
-# ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-# GATHER CONTEXT FOR JUDGE
-# ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-SPEC_FILE="$INCREMENT_DIR/spec.md"
-TASKS_FILE="$INCREMENT_DIR/tasks.md"
-
-if [ ! -f "$SPEC_FILE" ]; then
-    echo "❌ spec.md not found" >&2
-    exit 1
-fi
-
-if [ ! -f "$TASKS_FILE" ]; then
-    echo "❌ tasks.md not found" >&2
-    exit 1
-fi
-
-# Extract acceptance criteria from spec.md
-ACS=$(grep -E "^- \[.\] \*\*AC-" "$SPEC_FILE" 2>/dev/null || echo "")
-TOTAL_ACS=$(echo "$ACS" | wc -l | tr -d ' ')
-COMPLETED_ACS=$(echo "$ACS" | grep -c "^- \[x\]" || echo "0")
-
-# Extract tasks status
-TASKS=$(grep -E "^### T-[0-9]+" "$TASKS_FILE" 2>/dev/null || echo "")
-TOTAL_TASKS=$(echo "$TASKS" | wc -l | tr -d ' ')
-
-TASK_STATUS=$(grep -E "\*\*Status\*\*:" "$TASKS_FILE" 2>/dev/null || echo "")
-COMPLETED_TASKS=$(echo "$TASK_STATUS" | grep -c "\[x\] completed" || echo "0")
-
-# Get recent changes from git
-RECENT_CHANGES=$(git diff HEAD~5..HEAD --stat 2>/dev/null | head -20 || echo "No git history available")
-
-# Get test results from transcript
-TEST_RESULTS=$(tail -500 "$TRANSCRIPT_PATH" 2>/dev/null | grep -E "(PASS|FAIL|passed|failed|✓|✗)" | tail -20 || echo "No test results in transcript")
-
-# ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-# BUILD JUDGE PROMPT
-# ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-JUDGE_PROMPT=$(cat <<EOF
-You are a strict quality judge assessing whether work on increment $INCREMENT_ID is COMPLETE and PRODUCTION-READY.
-
-# Context
-
-## Acceptance Criteria Status
-- Total ACs: $TOTAL_ACS
-- Completed ACs: $COMPLETED_ACS
-
-## Tasks Status
-- Total Tasks: $TOTAL_TASKS
-- Completed Tasks: $COMPLETED_TASKS
-
-## Recent Changes
-\`\`\`
-$RECENT_CHANGES
-\`\`\`
-
-## Test Results
-\`\`\`
-$TEST_RESULTS
-\`\`\`
-
-## Specification
-$(head -100 "$SPEC_FILE")
-
-## Task Details
-$(head -100 "$TASKS_FILE")
-
-# Your Role
-
-You are the FINAL quality gate. Your job is to determine if this work is:
-1. **COMPLETE**: All acceptance criteria satisfied
-2. **TESTED**: Appropriate test coverage (unit + integration/E2E where needed)
-3. **PRODUCTION-READY**: No obvious bugs, security issues, or incomplete implementations
-
-# Evaluation Criteria
-
-✅ **APPROVE** if:
-- All ACs marked [x] AND actually implemented
-- Tests exist AND pass for critical paths
-- Code quality is production-ready
-- No half-finished implementations
-- Security best practices followed
-
-❌ **REJECT** if:
-- ACs marked [x] but not actually implemented (self-deception)
-- Missing tests for critical functionality
-- Obvious bugs or security vulnerabilities
-- Incomplete error handling
-- Poor code quality (will cause maintenance issues)
-
-# Output Format
-
-Respond with JSON ONLY:
-
-\`\`\`json
-{
-  "decision": "approve" | "reject",
-  "confidence": 0.0-1.0,
-  "reasoning": "Brief explanation (2-3 sentences)",
-  "concerns": ["concern 1", "concern 2"],
-  "recommendations": ["recommendation 1"]
-}
-\`\`\`
-
-**Be strict.** False approvals lead to production bugs. When in doubt, REJECT with clear guidance.
-EOF
-)
-
-# ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-# CALL CLAUDE API FOR JUDGMENT
-# ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-echo "🤖 Requesting LLM Judge assessment..." >&2
-
-# Check if API key exists
-if [ -z "$ANTHROPIC_API_KEY" ]; then
-    echo "⚠️  ANTHROPIC_API_KEY not set - skipping LLM judge (assuming approval)" >&2
-    exit 0
-fi
-
-# Create temporary file for request
-TMP_REQUEST=$(mktemp)
-cat > "$TMP_REQUEST" <<EOF
-{
-  "model": "claude-sonnet-4-20250514",
-  "max_tokens": 1024,
-  "messages": [
-    {
-      "role": "user",
-      "content": $(echo "$JUDGE_PROMPT" | jq -Rs .)
-    }
-  ]
-}
-EOF
-
-# Call Claude API
-RESPONSE=$(curl -s -X POST https://api.anthropic.com/v1/messages \
-    -H "Content-Type: application/json" \
-    -H "X-API-Key: $ANTHROPIC_API_KEY" \
-    -H "anthropic-version: 2023-06-01" \
-    -d @"$TMP_REQUEST")
-
-rm -f "$TMP_REQUEST"
-
-# Extract text content
-JUDGE_OUTPUT=$(echo "$RESPONSE" | jq -r '.content[0].text // ""')
-
-if [ -z "$JUDGE_OUTPUT" ]; then
-    echo "❌ Failed to get judge response" >&2
-    echo "API Response: $RESPONSE" >&2
-    exit 1
-fi
-
-# Parse JSON from output (extract JSON block)
-JUDGE_JSON=$(echo "$JUDGE_OUTPUT" | grep -A 100 '```json' | grep -B 100 '```' | grep -v '```' | jq -c '.')
-
-if [ -z "$JUDGE_JSON" ] || [ "$JUDGE_JSON" = "null" ]; then
-    echo "⚠️  Could not parse judge JSON - raw output:" >&2
-    echo "$JUDGE_OUTPUT" >&2
-    exit 1
-fi
-
-# Extract decision
-DECISION=$(echo "$JUDGE_JSON" | jq -r '.decision')
-CONFIDENCE=$(echo "$JUDGE_JSON" | jq -r '.confidence')
-REASONING=$(echo "$JUDGE_JSON" | jq -r '.reasoning')
-CONCERNS=$(echo "$JUDGE_JSON" | jq -r '.concerns | join(", ")')
-
-echo "" >&2
-echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━" >&2
-echo "🤖 LLM JUDGE ASSESSMENT" >&2
-echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━" >&2
-echo "" >&2
-echo "Decision: $DECISION (confidence: $CONFIDENCE)" >&2
-echo "Reasoning: $REASONING" >&2
-
-if [ "$DECISION" = "reject" ]; then
-    echo "" >&2
-    echo "❌ CONCERNS:" >&2
-    echo "$CONCERNS" >&2
-    echo "" >&2
-    echo "The LLM judge has REJECTED this work as not production-ready." >&2
-    echo "Address the concerns above and try again." >&2
-    echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━" >&2
-    exit 1
-fi
-
-echo "" >&2
-echo "✅ LLM judge APPROVED - work is production-ready" >&2
-echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━" >&2
-
-exit 0

package/plugins/specweave/hooks/log-decision.sh

@@ -1,168 +0,0 @@
-#!/bin/bash
-# log-decision.sh - Structured Decision Logging Utility
-#
-# Shared utility for all SpecWeave hooks to log decisions in structured JSON format.
-# Inspired by Claude Code 2.1.27's "Added tool call failures and denials to debug logs"
-#
-# Usage:
-#   source log-decision.sh
-#   log_decision "hook_name" "decision" "reason_code" "reason" "context_json" duration_ms
-#
-# Output:
-#   - Appends JSON entry to .specweave/logs/decisions.jsonl
-#   - Debug output to stderr when SPECWEAVE_DEBUG_HOOKS=1
-#
-# Schema:
-#   {
-#     "timestamp": "ISO8601",
-#     "hook": "stop-auto|stop-reflect|user-prompt-submit",
-#     "decision": "approve|block",
-#     "reason": "Human-readable reason",
-#     "reasonCode": "machine_parseable_enum",
-#     "durationMs": 123,
-#     "context": { ... hook-specific context ... }
-#   }
-
-# Max log size before rotation (10MB)
-_LOG_MAX_SIZE=10485760
-
-# Debug logging to stderr (only when SPECWEAVE_DEBUG_HOOKS=1)
-_log_debug() {
-    if [ "${SPECWEAVE_DEBUG_HOOKS:-0}" = "1" ]; then
-        local color_reset="\033[0m"
-        local color_cyan="\033[36m"
-        local color_green="\033[32m"
-        local color_red="\033[31m"
-        local color_yellow="\033[33m"
-
-        local level="$1"
-        shift
-        local message="$*"
-
-        local color="$color_cyan"
-        case "$level" in
-            "DECISION") color="$color_green" ;;
-            "BLOCK") color="$color_red" ;;
-            "WARN") color="$color_yellow" ;;
-        esac
-
-        echo -e "${color}[$level]${color_reset} $message" >&2
-    fi
-}
-
-# Rotate log file if it exceeds max size
-_rotate_log_if_needed() {
-    local log_file="$1"
-
-    if [ ! -f "$log_file" ]; then
-        return 0
-    fi
-
-    local file_size
-    # Cross-platform file size (macOS vs Linux)
-    if [[ "$OSTYPE" == "darwin"* ]]; then
-        file_size=$(stat -f%z "$log_file" 2>/dev/null || echo "0")
-    else
-        file_size=$(stat -c%s "$log_file" 2>/dev/null || echo "0")
-    fi
-
-    if [ "$file_size" -gt "$_LOG_MAX_SIZE" ]; then
-        _log_debug "WARN" "Log rotation triggered (${file_size} > ${_LOG_MAX_SIZE})"
-
-        # Keep last ~5MB by taking last N lines
-        # Estimate: average 500 bytes per line, so 5MB = ~10000 lines
-        local temp_file="${log_file}.tmp"
-        tail -n 10000 "$log_file" > "$temp_file" 2>/dev/null
-        mv "$temp_file" "$log_file" 2>/dev/null
-
-        _log_debug "INFO" "Log rotated, kept last 10000 entries"
-    fi
-}
-
-# Main logging function
-# Arguments:
-#   $1: hook_name - Name of the hook (stop-auto, stop-reflect, etc.)
-#   $2: decision - Decision made (approve, block)
-#   $3: reason_code - Machine-parseable reason code (session_inactive, work_remaining, etc.)
-#   $4: reason - Human-readable reason message
-#   $5: context_json - JSON string with hook-specific context (default: "{}")
-#   $6: duration_ms - Execution time in milliseconds (default: 0)
-log_decision() {
-    local hook_name="$1"
-    local decision="$2"
-    local reason_code="$3"
-    local reason="$4"
-    local context_json="${5:-"{}"}"
-    local duration_ms="${6:-0}"
-
-    # Compute paths based on current PROJECT_ROOT
-    local project_root="${PROJECT_ROOT:-$(pwd)}"
-    local log_dir="$project_root/.specweave/logs"
-    local log_file="$log_dir/decisions.jsonl"
-
-    # Validate required arguments
-    if [ -z "$hook_name" ] || [ -z "$decision" ] || [ -z "$reason_code" ]; then
-        _log_debug "WARN" "log_decision called with missing arguments"
-        return 1
-    fi
-
-    # Guard: only create dirs in initialized SpecWeave projects (prevents .specweave pollution)
-    if [ ! -d "$project_root/.specweave/increments" ] && [ ! -f "$project_root/.specweave/config.json" ]; then
-        _log_debug "WARN" "Not a SpecWeave project, skipping decision log"
-        return 0
-    fi
-
-    # Ensure logs directory exists
-    if [ ! -d "$log_dir" ]; then
-        mkdir -p "$log_dir" 2>/dev/null
-        _log_debug "INFO" "Created logs directory: $log_dir"
-    fi
-
-    # Check for log rotation
-    _rotate_log_if_needed "$log_file"
-
-    # Generate ISO 8601 timestamp
-    local timestamp
-    timestamp=$(date -u +%Y-%m-%dT%H:%M:%SZ 2>/dev/null || date +%Y-%m-%dT%H:%M:%SZ)
-
-    # Escape special characters in reason for JSON
-    local escaped_reason
-    escaped_reason=$(printf '%s' "$reason" | sed 's/\\/\\\\/g; s/"/\\"/g; s/	/\\t/g' | tr '\n' ' ')
-
-    # Build JSON entry using jq
-    local json_entry
-    json_entry=$(jq -n -c \
-        --arg timestamp "$timestamp" \
-        --arg hook "$hook_name" \
-        --arg decision "$decision" \
-        --arg reason "$escaped_reason" \
-        --arg reasonCode "$reason_code" \
-        --argjson durationMs "$duration_ms" \
-        --argjson context "$context_json" \
-        '{
-            timestamp: $timestamp,
-            hook: $hook,
-            decision: $decision,
-            reason: $reason,
-            reasonCode: $reasonCode,
-            durationMs: $durationMs,
-            context: $context
-        }' 2>/dev/null)
-
-    # Fallback if jq fails
-    if [ -z "$json_entry" ]; then
-        json_entry="{\"timestamp\":\"$timestamp\",\"hook\":\"$hook_name\",\"decision\":\"$decision\",\"reason\":\"$escaped_reason\",\"reasonCode\":\"$reason_code\",\"durationMs\":$duration_ms,\"context\":$context_json}"
-    fi
-
-    # Append to log file
-    printf '%s\n' "$json_entry" >> "$log_file" 2>/dev/null
-
-    # Debug output
-    if [ "$decision" = "block" ]; then
-        _log_debug "BLOCK" "$hook_name: $reason_code - $reason"
-    else
-        _log_debug "DECISION" "$hook_name: $decision ($reason_code)"
-    fi
-
-    return 0
-}

package/plugins/specweave/hooks/pre-compact.sh

@@ -1,64 +0,0 @@
-#!/bin/bash
-# pre-compact.sh - Context pressure signal handler (v1.0.262)
-#
-# Fires when Claude Code approaches context limits (before compaction).
-# Writes pressure state for UserPromptSubmit hook to read and reduce budget.
-#
-# Escalation: 1st → "elevated", 2nd → "critical", 3+ → "emergency"
-set +e
-
-[[ "${SPECWEAVE_DISABLE_HOOKS:-0}" == "1" ]] && echo '{"continue":true}' && exit 0
-
-# Find project root
-PROJECT_ROOT="$PWD"
-while [[ "$PROJECT_ROOT" != "/" ]] && [[ ! -f "$PROJECT_ROOT/.specweave/config.json" ]]; do
-  PROJECT_ROOT=$(dirname "$PROJECT_ROOT")
-done
-[[ ! -f "$PROJECT_ROOT/.specweave/config.json" ]] && echo '{"continue":true}' && exit 0
-
-STATE_DIR="$PROJECT_ROOT/.specweave/state"
-PRESSURE_FILE="$STATE_DIR/context-pressure.json"
-mkdir -p "$STATE_DIR" 2>/dev/null
-
-# Read previous compaction count for escalation
-PREV_COUNT=0
-if [[ -f "$PRESSURE_FILE" ]] && command -v jq >/dev/null 2>&1; then
-  PREV_COUNT=$(jq -r '.compactionCount // 0' "$PRESSURE_FILE" 2>/dev/null || echo "0")
-fi
-NEW_COUNT=$((PREV_COUNT + 1))
-
-# Determine level: 1 = elevated, 2 = critical, 3+ = emergency
-LEVEL="elevated"
-[[ "$NEW_COUNT" -ge 2 ]] && LEVEL="critical"
-[[ "$NEW_COUNT" -ge 3 ]] && LEVEL="emergency"
-
-cat > "$PRESSURE_FILE" <<EOF
-{"level":"$LEVEL","compactionCount":$NEW_COUNT,"lastCompaction":"$(date -Iseconds)"}
-EOF
-
-# Write prompt-health-alert.json with remediation advice (v1.0.268)
-ADVICE=""
-case "$LEVEL" in
-  elevated)
-    ADVICE="Context budget reduced to compact. Session is running long."
-    ;;
-  critical)
-    ADVICE="Context budget at minimal. Consider starting a new session or setting contextBudget.level to off."
-    ;;
-  emergency)
-    ADVICE="Context budget stripped to off (3+ compactions). Strongly recommend starting a new session."
-    ;;
-esac
-
-ALERT_FILE="$STATE_DIR/prompt-health-alert.json"
-cat > "$ALERT_FILE" <<EOF
-{"level":"$LEVEL","compactionCount":$NEW_COUNT,"advice":"$ADVICE","timestamp":"$(date -Iseconds)"}
-EOF
-
-# Log compaction event for dashboard activity stream
-_LOG_DIR="$PROJECT_ROOT/.specweave/logs"
-mkdir -p "$_LOG_DIR" 2>/dev/null
-echo "[$(date -Iseconds)] COMPACTION #$NEW_COUNT | level=$LEVEL | $ADVICE" >> "$_LOG_DIR/prompt-health.log" 2>/dev/null
-
-echo '{"continue":true}'
-exit 0

package/plugins/specweave/hooks/startup-health-check.sh

@@ -1,64 +0,0 @@
-#!/bin/bash
-
-# ╔══════════════════════════════════════════════════════════════════════════╗
-# ║  SpecWeave Startup Health Check                                           ║
-# ║  Runs on Claude Code startup to detect and auto-fix plugin issues         ║
-# ║  This hook runs BEFORE plugin loading, catching issues early              ║
-# ╚══════════════════════════════════════════════════════════════════════════╝
-
-# Detect project root via walk-up
-_DIR="$PWD"
-_LIMIT=0
-while [ "$_DIR" != "/" ] && [ ! -f "$_DIR/.specweave/config.json" ] && [ $_LIMIT -lt 50 ]; do
-  _DIR=$(dirname "$_DIR")
-  _LIMIT=$((_LIMIT + 1))
-done
-if [ ! -f "$_DIR/.specweave/config.json" ]; then
-    exit 0
-fi
-PROJECT_ROOT="$_DIR"
-
-HEALTH_LOG="$HOME/.claude/plugins/.health-check.log"
-
-# Function to log with timestamp
-log() {
-    echo "[$(date '+%Y-%m-%d %H:%M:%S')] $1" >> "$HEALTH_LOG"
-}
-
-# Rotate log if too large (>100KB)
-if [ -f "$HEALTH_LOG" ] && [ $(stat -f%z "$HEALTH_LOG" 2>/dev/null || stat -c%s "$HEALTH_LOG" 2>/dev/null) -gt 102400 ]; then
-    mv "$HEALTH_LOG" "${HEALTH_LOG}.old"
-fi
-
-log "=== Health check started ==="
-
-# ============================================================================
-# SESSION CLEANUP: ALWAYS clear auto-mode.json (CRITICAL for session-scoped auto)
-# Auto mode MUST be session-scoped - each new Claude Code session starts FRESH.
-# The file is per-project but MUST only affect the session that created it.
-# ============================================================================
-
-AUTO_MODE_FILE="$PROJECT_ROOT/.specweave/state/auto-mode.json"
-STATE_DIR="$PROJECT_ROOT/.specweave/state"
-
-# ALWAYS clear auto-mode.json on session start
-# This ensures auto mode is truly session-scoped:
-# - Session A runs /sw:auto → creates auto-mode.json
-# - Session A closes → file remains but is now orphaned
-# - Session B starts → SessionStart hook clears the file
-# - Session B is NOT in auto mode (correct behavior)
-if [ -f "$AUTO_MODE_FILE" ]; then
-    log "SESSION START: Clearing previous auto-mode.json (session-scoped cleanup)"
-    rm -f "$AUTO_MODE_FILE" 2>/dev/null
-    rm -f "$STATE_DIR/.stop-auto-dedup" 2>/dev/null
-    rm -f "$STATE_DIR/.stop-auto-dedup-prev" 2>/dev/null
-    rm -f "$STATE_DIR/.stop-auto-turns" 2>/dev/null
-    log "Cleared all auto-mode session files - new session starts fresh"
-fi
-
-# v1.0.344 (0394): Removed marketplace directory checks (Checks 1-4).
-# Plugins are now installed on-demand via `npx vskill install --repo` in user-prompt-submit.sh.
-# No local marketplace directory at ~/.claude/plugins/marketplaces/specweave is required.
-
-log "OK: Health check passed (session cleanup complete)"
-exit 0