devforgeai 1.0.5 → 1.0.7

package/src/claude/scripts/requirements.txt

@@ -0,0 +1,8 @@
+# DevForgeAI CLI Requirements
+
+# Core dependencies
+PyYAML>=6.0
+
+# Development dependencies (optional)
+pytest>=7.0  # For running tests
+pytest-cov>=4.0  # For coverage reports

package/src/claude/scripts/session_catalog.sh

@@ -0,0 +1,543 @@
+#!/bin/bash
+#
+# Session Catalog Functions for STORY-223
+# Provides session file cataloging, dependency graph building, and session chain tracking
+#
+# Usage: source .claude/scripts/session_catalog.sh
+#
+# Functions:
+#   catalog_session_files(directory)    - Map plans to stories to artifacts (AC#1)
+#   build_dependency_graph(directory)   - Build file dependency graph (AC#2)
+#   track_session_chains(directory)     - Track session continuity chains (AC#3)
+#
+# Refactored: Extracted helper functions, reduced complexity, improved performance
+#
+
+# ============================================================================
+# Helper Functions (DRY - extracted repeated patterns)
+# ============================================================================
+
+# Find files matching a pattern and store in array variable
+# Usage: find_files_into_array "/path" "*.md" result_array
+find_files_into_array() {
+    local directory="$1"
+    local pattern="$2"
+    local -n _result_array="$3"
+    _result_array=()
+    while IFS= read -r -d '' file; do
+        _result_array+=("$file")
+    done < <(find "$directory" -name "$pattern" -type f -print0 2>/dev/null)
+}
+
+# Classify file type based on path
+# Usage: file_type=$(classify_file_type "/path/to/file")
+classify_file_type() {
+    local file="$1"
+    case "$file" in
+        */plans/*) echo "plan" ;;
+        */sessions/*) echo "session" ;;
+        */artifacts/*) echo "artifact" ;;
+        */Stories/*|*/stories/*) echo "story" ;;
+        *) echo "other" ;;
+    esac
+}
+
+# Build JSON array from bash array
+# Usage: json=$(build_json_array "${items[@]}")
+build_json_array() {
+    local items=("$@")
+    local result=""
+    for item in "${items[@]}"; do
+        # Escape special JSON characters
+        local escaped="${item//\\/\\\\}"
+        escaped="${escaped//\"/\\\"}"
+        [[ -n "$result" ]] && result+=","
+        result+="\"$escaped\""
+    done
+    echo "[$result]"
+}
+
+# ============================================================================
+# AC#1: Catalog Session Files
+# Maps plan files -> story references -> associated artifacts
+# ============================================================================
+
+# Extract story references from plan file YAML frontmatter
+# Usage: extract_story_refs "/path/to/plan.md" result_array
+_extract_story_refs() {
+    local plan_file="$1"
+    local -n _refs="$2"
+    _refs=()
+    while IFS= read -r line; do
+        if [[ "$line" =~ ^[[:space:]]*-[[:space:]]*(STORY-[0-9]+) ]]; then
+            _refs+=("${BASH_REMATCH[1]}")
+        fi
+    done < <(sed -n '/^related_stories:/,/^[^[:space:]-]/p' "$plan_file" 2>/dev/null)
+}
+
+# Process plan files and populate mappings
+# Uses associative arrays for O(1) lookups, then converts to JSON at end
+_process_plan_files() {
+    local directory="$1"
+    local -n _p2s_map="$2"
+    local -n _s2p_map="$3"
+    local -n _files_json="$4"
+
+    local plan_files=()
+    while IFS= read -r -d '' file; do
+        plan_files+=("$file")
+    done < <(find "$directory/plans" -name "*.md" -type f -print0 2>/dev/null)
+
+    for plan_file in "${plan_files[@]}"; do
+        local plan_id
+        plan_id=$(basename "$plan_file" .md)
+
+        local story_refs=()
+        _extract_story_refs "$plan_file" story_refs
+
+        if [[ ${#story_refs[@]} -gt 0 ]]; then
+            # Store as comma-separated for later conversion
+            _p2s_map["$plan_id"]="${story_refs[*]}"
+
+            # Build reverse mapping
+            for story in "${story_refs[@]}"; do
+                if [[ -n "${_s2p_map[$story]:-}" ]]; then
+                    _s2p_map["$story"]+=" $plan_id"
+                else
+                    _s2p_map["$story"]="$plan_id"
+                fi
+            done
+        fi
+
+        # Collect file entry
+        _files_json+=("{\"path\":\"$plan_file\",\"type\":\"plan\"}")
+    done
+}
+
+# Process artifact directories and populate mappings
+_process_artifacts() {
+    local directory="$1"
+    local -n _s2a_map="$2"
+    local -n _files_json="$3"
+
+    local artifact_dirs=()
+    while IFS= read -r -d '' dir; do
+        artifact_dirs+=("$dir")
+    done < <(find "$directory/artifacts" -mindepth 1 -maxdepth 1 -type d -print0 2>/dev/null)
+
+    for artifact_dir in "${artifact_dirs[@]}"; do
+        local story_id
+        story_id=$(basename "$artifact_dir")
+        local artifact_names=()
+
+        while IFS= read -r -d '' file; do
+            local filename
+            filename=$(basename "$file")
+            artifact_names+=("$filename")
+            _files_json+=("{\"path\":\"$file\",\"type\":\"artifact\"}")
+        done < <(find "$artifact_dir" -type f -print0 2>/dev/null)
+
+        if [[ ${#artifact_names[@]} -gt 0 ]]; then
+            _s2a_map["$story_id"]="${artifact_names[*]}"
+        fi
+    done
+}
+
+# Convert associative array to JSON object with array values
+# Usage: _assoc_to_json_object assoc_array -> {"key":["val1","val2"]}
+_assoc_to_json_object() {
+    local -n _assoc="$1"
+    local result="{"
+    local first=true
+
+    for key in "${!_assoc[@]}"; do
+        [[ "$first" == "true" ]] || result+=","
+        first=false
+
+        # Convert space-separated values to JSON array
+        local vals="${_assoc[$key]}"
+        local arr_json=""
+        for val in $vals; do
+            [[ -n "$arr_json" ]] && arr_json+=","
+            arr_json+="\"$val\""
+        done
+        result+="\"$key\":[$arr_json]"
+    done
+
+    echo "$result}"
+}
+
+catalog_session_files() {
+    local directory="${1:-.}"
+
+    # Use associative arrays for O(1) lookups during processing
+    declare -A plan_to_story_map
+    declare -A story_to_artifacts_map
+    declare -A story_to_plans_map
+    local files_entries=()
+
+    # Process all plan files
+    _process_plan_files "$directory" plan_to_story_map story_to_plans_map files_entries
+
+    # Process all artifacts
+    _process_artifacts "$directory" story_to_artifacts_map files_entries
+
+    # Process session files
+    while IFS= read -r -d '' file; do
+        files_entries+=("{\"path\":\"$file\",\"type\":\"session\"}")
+    done < <(find "$directory/sessions" -name "*.json" -type f -print0 2>/dev/null)
+
+    # Convert to JSON (single jq call for final assembly)
+    local p2s_json s2a_json s2p_json files_json
+    p2s_json=$(_assoc_to_json_object plan_to_story_map)
+    s2a_json=$(_assoc_to_json_object story_to_artifacts_map)
+    s2p_json=$(_assoc_to_json_object story_to_plans_map)
+
+    # Build files array
+    if [[ ${#files_entries[@]} -eq 0 ]]; then
+        files_json="[]"
+    else
+        files_json="[$(IFS=,; echo "${files_entries[*]}")]"
+    fi
+
+    # Single jq call to assemble final output (performance optimization)
+    jq -n -c \
+        --argjson plan_to_story "$p2s_json" \
+        --argjson story_to_artifacts "$s2a_json" \
+        --argjson story_to_plans "$s2p_json" \
+        --argjson files "$files_json" \
+        '{
+            "plan_to_story_map": $plan_to_story,
+            "story_to_artifacts_map": $story_to_artifacts,
+            "story_to_plans_map": $story_to_plans,
+            "files": $files
+        }'
+}
+
+# ============================================================================
+# AC#2: Build Dependency Graph
+# Analyzes file references and builds dependency graph
+# Performance: O(n*m) where n=files, m=average file size for grep
+# Optimized: Uses associative arrays for O(1) lookups in cycle detection
+# ============================================================================
+
+# Classify dependency type based on source and target paths
+# Usage: dep_type=$(_classify_dependency_type "/source/path" "/target/path")
+_classify_dependency_type() {
+    local source_path="$1"
+    local target_path="$2"
+
+    if [[ "$source_path" == */plans/* && "$target_path" == */Stories/* ]]; then
+        echo "plan_to_story"
+    elif [[ "$source_path" == */sessions/* && "$target_path" == */sessions/* ]]; then
+        echo "session_chain"
+    else
+        echo "reference"
+    fi
+}
+
+# Build nodes map from file list (batch operation)
+# Usage: nodes_json=$(_build_nodes_map "${files[@]}")
+_build_nodes_map() {
+    local files=("$@")
+    local result="{"
+    local first=true
+
+    for file in "${files[@]}"; do
+        local file_id file_type
+        file_id=$(basename "$file")
+        file_type=$(classify_file_type "$file")
+
+        [[ "$first" == "true" ]] || result+=","
+        first=false
+        result+="\"$file_id\":{\"path\":\"$file\",\"type\":\"$file_type\"}"
+    done
+
+    echo "$result}"
+}
+
+# Find file references using optimized grep (single pass per file)
+# Populates associative array with source|target pairs
+_find_file_references() {
+    local -n _dep_entries="$1"
+    shift
+    local files=("$@")
+
+    # Build lookup sets for O(1) target matching
+    declare -A file_ids
+    declare -A file_basenames
+    declare -A file_paths
+    for file in "${files[@]}"; do
+        local fid fbase
+        fid=$(basename "$file")
+        fbase="${fid%.*}"
+        file_ids["$fid"]=1
+        file_basenames["$fbase"]="$fid"
+        file_paths["$fid"]="$file"
+    done
+
+    # Process each source file
+    for source_file in "${files[@]}"; do
+        local source_id
+        source_id=$(basename "$source_file")
+
+        # Single grep to get all potential references
+        local file_content
+        file_content=$(cat "$source_file" 2>/dev/null) || continue
+
+        # Check each potential target
+        for target_id in "${!file_ids[@]}"; do
+            [[ "$source_id" == "$target_id" ]] && continue
+
+            local target_basename="${target_id%.*}"
+
+            # Check if file content references target
+            if [[ "$file_content" == *"$target_id"* ]] || [[ "$file_content" == *"$target_basename"* ]]; then
+                local target_file="${file_paths[$target_id]}"
+                local dep_type
+                dep_type=$(_classify_dependency_type "$source_file" "$target_file")
+                _dep_entries+=("{\"source\":\"$source_id\",\"target\":\"$target_id\",\"type\":\"$dep_type\"}")
+            fi
+        done
+    done
+}
+
+# Detect circular dependencies using associative array for O(n) performance
+# Usage: circular_json=$(_detect_circular_deps "${dep_entries[@]}")
+_detect_circular_deps() {
+    local dep_entries=("$@")
+
+    # Build edge set for O(1) reverse lookup
+    declare -A edge_set
+    declare -A seen_cycles
+    local circular_entries=()
+
+    for entry in "${dep_entries[@]}"; do
+        # Extract source and target from JSON entry
+        local source target
+        source=$(echo "$entry" | grep -o '"source":"[^"]*"' | cut -d'"' -f4)
+        target=$(echo "$entry" | grep -o '"target":"[^"]*"' | cut -d'"' -f4)
+        edge_set["$source|$target"]=1
+    done
+
+    # Check for reverse edges (cycles)
+    for edge in "${!edge_set[@]}"; do
+        IFS='|' read -r source target <<< "$edge"
+        local reverse_edge="$target|$source"
+
+        if [[ -n "${edge_set[$reverse_edge]:-}" ]]; then
+            # Normalize cycle representation (alphabetical order)
+            local cycle_key
+            if [[ "$source" < "$target" ]]; then
+                cycle_key="$source|$target"
+            else
+                cycle_key="$target|$source"
+            fi
+
+            # Only add if not already seen
+            if [[ -z "${seen_cycles[$cycle_key]:-}" ]]; then
+                seen_cycles["$cycle_key"]=1
+                if [[ "$source" < "$target" ]]; then
+                    circular_entries+=("[\"$source\",\"$target\"]")
+                else
+                    circular_entries+=("[\"$target\",\"$source\"]")
+                fi
+            fi
+        fi
+    done
+
+    # Build JSON array
+    if [[ ${#circular_entries[@]} -eq 0 ]]; then
+        echo "[]"
+    else
+        echo "[$(IFS=,; echo "${circular_entries[*]}")]"
+    fi
+}
+
+build_dependency_graph() {
+    local directory="${1:-.}"
+
+    # Find all relevant files
+    local all_files=()
+    while IFS= read -r -d '' file; do
+        all_files+=("$file")
+    done < <(find "$directory" \( -name "*.md" -o -name "*.json" -o -name "*.sh" \) -type f -print0 2>/dev/null)
+
+    # Build nodes map (batch operation - no jq calls in loop)
+    local nodes_json
+    nodes_json=$(_build_nodes_map "${all_files[@]}")
+
+    # Find all dependencies
+    local dep_entries=()
+    _find_file_references dep_entries "${all_files[@]}"
+
+    # Detect circular dependencies (O(n) with hash lookup)
+    local circular_json
+    circular_json=$(_detect_circular_deps "${dep_entries[@]}")
+
+    # Build dependencies JSON array
+    local deps_json
+    if [[ ${#dep_entries[@]} -eq 0 ]]; then
+        deps_json="[]"
+    else
+        deps_json="[$(IFS=,; echo "${dep_entries[*]}")]"
+    fi
+
+    # Single jq call for final assembly
+    jq -n -c \
+        --argjson deps "$deps_json" \
+        --argjson nodes "$nodes_json" \
+        --argjson circular "$circular_json" \
+        '{
+            "dependencies": $deps,
+            "nodes": $nodes,
+            "circular_dependencies": $circular
+        }'
+}
+
+# ============================================================================
+# AC#3: Track Session Chains
+# Identifies parent-child relationships via parentUuid
+# Optimized: Reduced jq calls, uses pure bash for chain building
+# ============================================================================
+
+# Extract uuid and parentUuid from session file (single jq call)
+# Usage: read uuid parent_uuid < <(_extract_session_uuids "$file")
+_extract_session_uuids() {
+    local file="$1"
+    jq -r '[.uuid // "", .parentUuid // ""] | @tsv' "$file" 2>/dev/null
+}
+
+# Build session chain using BFS (pure bash, no jq in loop)
+# Usage: chain_json=$(_build_session_chain "$root_uuid" children_map visited_set)
+_build_session_chain() {
+    local root_uuid="$1"
+    local -n _children="$2"
+    local -n _visited="$3"
+
+    local depth=0
+    local node_list=("$root_uuid")
+    local queue=("$root_uuid")
+    _visited["$root_uuid"]=1
+
+    while [[ ${#queue[@]} -gt 0 ]]; do
+        local current="${queue[0]}"
+        queue=("${queue[@]:1}")
+
+        local children="${_children[$current]:-}"
+        for child in $children; do
+            if [[ -z "${_visited[$child]:-}" ]]; then
+                _visited["$child"]=1
+                queue+=("$child")
+                node_list+=("$child")
+                ((depth++))
+            fi
+        done
+    done
+
+    # Build JSON without jq
+    local nodes_json
+    nodes_json=$(build_json_array "${node_list[@]}")
+    echo "{\"root\":\"$root_uuid\",\"nodes\":$nodes_json,\"depth\":$depth}"
+}
+
+track_session_chains() {
+    local session_directory="${1:-.}"
+
+    # Find all session files
+    local session_files=()
+    while IFS= read -r -d '' file; do
+        session_files+=("$file")
+    done < <(find "$session_directory" -name "*.json" -type f -print0 2>/dev/null)
+
+    # Build session maps: uuid -> file, uuid -> parent, parent -> children
+    declare -A session_map
+    declare -A parent_map
+    declare -A children_map
+
+    for file in "${session_files[@]}"; do
+        local uuid parent_uuid
+        read -r uuid parent_uuid < <(_extract_session_uuids "$file")
+
+        [[ -z "$uuid" ]] && continue
+
+        session_map["$uuid"]="$file"
+
+        if [[ -n "$parent_uuid" ]]; then
+            parent_map["$uuid"]="$parent_uuid"
+            children_map["$parent_uuid"]+="$uuid "
+        fi
+    done
+
+    # Categorize sessions: roots (no parent) and orphans (missing parent)
+    local root_uuids=()
+    local orphan_uuids=()
+
+    for uuid in "${!session_map[@]}"; do
+        local parent="${parent_map[$uuid]:-}"
+
+        if [[ -z "$parent" ]]; then
+            root_uuids+=("$uuid")
+        elif [[ -z "${session_map[$parent]:-}" ]]; then
+            orphan_uuids+=("$uuid")
+        fi
+    done
+
+    # Build chains from each root (pure bash)
+    local chain_entries=()
+    for root in "${root_uuids[@]}"; do
+        declare -A visited
+        local chain_json
+        chain_json=$(_build_session_chain "$root" children_map visited)
+        chain_entries+=("$chain_json")
+        unset visited
+    done
+
+    # Build output JSON arrays without intermediate jq calls
+    local roots_json orphans_json chains_json
+    roots_json=$(build_json_array "${root_uuids[@]}")
+    orphans_json=$(build_json_array "${orphan_uuids[@]}")
+
+    if [[ ${#chain_entries[@]} -eq 0 ]]; then
+        chains_json="[]"
+    else
+        chains_json="[$(IFS=,; echo "${chain_entries[*]}")]"
+    fi
+
+    # Single jq call for final assembly
+    jq -n -c \
+        --argjson chains "$chains_json" \
+        --argjson roots "$roots_json" \
+        --argjson orphans "$orphans_json" \
+        '{
+            "session_chains": $chains,
+            "root_sessions": $roots,
+            "orphan_sessions": $orphans
+        }'
+}
+
+# ============================================================================
+# Export functions
+# ============================================================================
+
+# Primary API functions
+export -f catalog_session_files
+export -f build_dependency_graph
+export -f track_session_chains
+
+# Helper functions (also exported for testing/reuse)
+export -f find_files_into_array
+export -f classify_file_type
+export -f build_json_array
+
+# Internal helper functions (exported for subshell compatibility)
+export -f _extract_story_refs
+export -f _process_plan_files
+export -f _process_artifacts
+export -f _assoc_to_json_object
+export -f _classify_dependency_type
+export -f _build_nodes_map
+export -f _find_file_references
+export -f _detect_circular_deps
+export -f _extract_session_uuids
+export -f _build_session_chain

package/src/claude/scripts/setup.py

@@ -0,0 +1,55 @@
+"""
+DevForgeAI CLI Setup
+
+Installation:
+    pip install -e .claude/scripts/
+
+Usage:
+    devforgeai validate-dod devforgeai/specs/Stories/STORY-001.story.md
+    devforgeai check-git
+    devforgeai validate-context
+"""
+
+from setuptools import setup, find_packages
+from pathlib import Path
+
+# Read version from __init__.py
+init_file = Path(__file__).parent / 'devforgeai_cli' / '__init__.py'
+version = '0.1.0'
+for line in init_file.read_text().splitlines():
+    if line.startswith('__version__'):
+        version = line.split('=')[1].strip(' \'"')
+        break
+
+setup(
+    name='devforgeai-cli',
+    version=version,
+    description='DevForgeAI workflow validators for spec-driven development',
+    long_description='Automated validation tools that prevent autonomous deferrals, '
+                     'validate Git availability, and enforce workflow quality gates.',
+    author='DevForgeAI Contributors',
+    python_requires='>=3.8',
+    packages=find_packages(),
+    install_requires=[
+        'PyYAML>=6.0',
+    ],
+    entry_points={
+        'console_scripts': [
+            # Renamed from 'devforgeai' to avoid conflict with npm global CLI
+            # The npm CLI (devforgeai) is user-facing for installation
+            # This Python CLI is internal for framework validation
+            'devforgeai-validate=devforgeai_cli.cli:main',
+        ],
+    },
+    classifiers=[
+        'Development Status :: 3 - Alpha',
+        'Intended Audience :: Developers',
+        'Topic :: Software Development :: Quality Assurance',
+        'Programming Language :: Python :: 3',
+        'Programming Language :: Python :: 3.8',
+        'Programming Language :: Python :: 3.9',
+        'Programming Language :: Python :: 3.10',
+        'Programming Language :: Python :: 3.11',
+        'Programming Language :: Python :: 3.12',
+    ],
+)

package/src/claude/scripts/start-devforgeai.sh

@@ -0,0 +1,16 @@
+#!/bin/bash
+# DevForgeAI Claude Launcher
+# Usage: ./start-devforgeai.sh [claude args...]
+#
+# Appends DevForgeAI core directives to Claude's system prompt
+
+SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
+SYSTEM_PROMPT_FILE="$SCRIPT_DIR/../system-prompt-core.md"
+
+if [ ! -f "$SYSTEM_PROMPT_FILE" ]; then
+    echo "Error: system-prompt-core.md not found at $SYSTEM_PROMPT_FILE"
+    exit 1
+fi
+
+# Launch Claude with appended system prompt
+claude --append-system-prompt "$(cat "$SYSTEM_PROMPT_FILE")" "$@"

package/src/claude/scripts/statusline.sh

@@ -0,0 +1,27 @@
+#!/bin/bash
+input=$(cat)
+
+MODEL=$(echo "$input" | jq -r '.model.display_name')
+DIR=$(echo "$input" | jq -r '.workspace.current_dir')
+COST=$(echo "$input" | jq -r '.cost.total_cost_usd // 0')
+PCT=$(echo "$input" | jq -r '.context_window.used_percentage // 0' | cut -d. -f1)
+DURATION_MS=$(echo "$input" | jq -r '.cost.total_duration_ms // 0')
+
+CYAN='\033[36m'; GREEN='\033[32m'; YELLOW='\033[33m'; RED='\033[31m'; RESET='\033[0m'
+
+# Pick bar color based on context usage
+if [ "$PCT" -ge 90 ]; then BAR_COLOR="$RED"
+elif [ "$PCT" -ge 70 ]; then BAR_COLOR="$YELLOW"
+else BAR_COLOR="$GREEN"; fi
+
+FILLED=$((PCT / 10)); EMPTY=$((10 - FILLED))
+BAR=$(printf "%${FILLED}s" | tr ' ' '█')$(printf "%${EMPTY}s" | tr ' ' '░')
+
+MINS=$((DURATION_MS / 60000)); SECS=$(((DURATION_MS % 60000) / 1000))
+
+BRANCH=""
+git rev-parse --git-dir > /dev/null 2>&1 && BRANCH=" | 🌿 $(git branch --show-current 2>/dev/null)"
+
+echo -e "${CYAN}[$MODEL]${RESET} 📁 ${DIR##*/}$BRANCH"
+COST_FMT=$(printf '$%.2f' "$COST")
+echo -e "${BAR_COLOR}${BAR}${RESET} ${PCT}% | ${YELLOW}${COST_FMT}${RESET} | ⏱️ ${MINS}m ${SECS}s"