autonomous-coding-toolkit 1.0.0

package/scripts/tests/test-agents-md.sh

@@ -0,0 +1,87 @@
+#!/usr/bin/env bash
+# Test AGENTS.md generation
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+source "$SCRIPT_DIR/../lib/run-plan-parser.sh"
+source "$SCRIPT_DIR/../lib/run-plan-prompt.sh"
+
+FAILURES=0
+TESTS=0
+
+assert_eq() {
+    local desc="$1" expected="$2" actual="$3"
+    TESTS=$((TESTS + 1))
+    if [[ "$expected" != "$actual" ]]; then
+        echo "FAIL: $desc"
+        echo "  expected: $expected"
+        echo "  actual:   $actual"
+        FAILURES=$((FAILURES + 1))
+    else
+        echo "PASS: $desc"
+    fi
+}
+
+assert_contains() {
+    local desc="$1" needle="$2" haystack="$3"
+    TESTS=$((TESTS + 1))
+    if echo "$haystack" | grep -qF "$needle"; then
+        echo "PASS: $desc"
+    else
+        echo "FAIL: $desc"
+        echo "  expected to contain: $needle"
+        FAILURES=$((FAILURES + 1))
+    fi
+}
+
+WORK=$(mktemp -d)
+trap 'rm -rf "$WORK"' EXIT
+
+# Create a test plan
+cat > "$WORK/plan.md" << 'PLAN'
+## Batch 1: Setup foundation
+- Create: `src/lib.py`
+- Test: `tests/test_lib.py`
+
+**Step 1:** Create files
+
+## Batch 2: Core Logic
+- Create: `src/core.py`
+- Modify: `src/lib.py`
+
+**Step 1:** Add core
+PLAN
+
+# Generate AGENTS.md
+generate_agents_md "$WORK/plan.md" "$WORK" "headless"
+
+# Test: file created
+TESTS=$((TESTS + 1))
+if [[ -f "$WORK/AGENTS.md" ]]; then
+    echo "PASS: AGENTS.md created"
+else
+    echo "FAIL: AGENTS.md should be created"
+    FAILURES=$((FAILURES + 1))
+fi
+
+output=$(cat "$WORK/AGENTS.md")
+assert_contains "has batch count" "2 batches" "$output"
+assert_contains "has mode" "headless" "$output"
+assert_contains "has tools" "Bash" "$output"
+assert_contains "has plan name" "plan.md" "$output"
+assert_contains "has batch 1 title" "Setup foundation" "$output"
+assert_contains "has batch 2 title" "Core Logic" "$output"
+assert_contains "has guidelines" "quality gate" "$output"
+
+# Test: team mode
+generate_agents_md "$WORK/plan.md" "$WORK" "team"
+output=$(cat "$WORK/AGENTS.md")
+assert_contains "team mode updated" "team" "$output"
+
+echo ""
+echo "Results: $((TESTS - FAILURES))/$TESTS passed"
+if [[ $FAILURES -gt 0 ]]; then
+    echo "FAILURES: $FAILURES"
+    exit 1
+fi
+echo "ALL PASSED"

package/scripts/tests/test-analyze-report.sh

@@ -0,0 +1,114 @@
+#!/usr/bin/env bash
+# Test analyze-report.sh — verifies refactored behavior using ollama.sh
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+AR="$SCRIPT_DIR/../analyze-report.sh"
+
+FAILURES=0
+TESTS=0
+
+assert_eq() {
+    local desc="$1" expected="$2" actual="$3"
+    TESTS=$((TESTS + 1))
+    if [[ "$expected" != "$actual" ]]; then
+        echo "FAIL: $desc"
+        echo "  expected: $expected"
+        echo "  actual:   $actual"
+        FAILURES=$((FAILURES + 1))
+    else
+        echo "PASS: $desc"
+    fi
+}
+
+assert_exit() {
+    local desc="$1" expected_exit="$2"
+    shift 2
+    local actual_exit=0
+    "$@" >/dev/null 2>&1 || actual_exit=$?
+    TESTS=$((TESTS + 1))
+    if [[ "$expected_exit" != "$actual_exit" ]]; then
+        echo "FAIL: $desc"
+        echo "  expected exit: $expected_exit"
+        echo "  actual exit:   $actual_exit"
+        FAILURES=$((FAILURES + 1))
+    else
+        echo "PASS: $desc"
+    fi
+}
+
+# === Sources shared libraries ===
+
+TESTS=$((TESTS + 1))
+if grep -q 'source.*lib/common.sh' "$AR"; then
+    echo "PASS: analyze-report.sh sources lib/common.sh"
+else
+    echo "FAIL: analyze-report.sh sources lib/common.sh"
+    FAILURES=$((FAILURES + 1))
+fi
+
+TESTS=$((TESTS + 1))
+if grep -q 'source.*lib/ollama.sh' "$AR"; then
+    echo "PASS: analyze-report.sh sources lib/ollama.sh"
+else
+    echo "FAIL: analyze-report.sh sources lib/ollama.sh"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Uses ollama_query ===
+
+TESTS=$((TESTS + 1))
+if grep -q 'ollama_query' "$AR"; then
+    echo "PASS: uses ollama_query function"
+else
+    echo "FAIL: should use ollama_query function"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Uses ollama_extract_json ===
+
+TESTS=$((TESTS + 1))
+if grep -q 'ollama_extract_json' "$AR"; then
+    echo "PASS: uses ollama_extract_json function"
+else
+    echo "FAIL: should use ollama_extract_json function"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === No inline curl to Ollama ===
+
+TESTS=$((TESTS + 1))
+if grep -q 'curl.*api/generate' "$AR"; then
+    echo "FAIL: still has inline curl to Ollama API"
+    FAILURES=$((FAILURES + 1))
+else
+    echo "PASS: no inline curl to Ollama API"
+fi
+
+# === CLI tests ===
+
+assert_exit "no args exits 1" 1 bash "$AR"
+assert_exit "nonexistent file exits 1" 1 bash "$AR" /nonexistent/report.md
+
+# === Dry run works ===
+
+WORK=$(mktemp -d)
+trap 'rm -rf "$WORK"' EXIT
+echo "# Test report" > "$WORK/report.md"
+output=$(bash "$AR" "$WORK/report.md" --dry-run 2>&1) || true
+TESTS=$((TESTS + 1))
+if echo "$output" | grep -q "DRY RUN"; then
+    echo "PASS: dry-run mode works"
+else
+    echo "FAIL: dry-run mode should output DRY RUN"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Summary ===
+echo ""
+echo "Results: $((TESTS - FAILURES))/$TESTS passed"
+if [[ $FAILURES -gt 0 ]]; then
+    echo "FAILURES: $FAILURES"
+    exit 1
+fi
+echo "ALL PASSED"

package/scripts/tests/test-architecture-map.sh

@@ -0,0 +1,89 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+source "$SCRIPT_DIR/test-helpers.sh"
+
+ARCH_MAP="$SCRIPT_DIR/../architecture-map.sh"
+
+# --- Test: --help exits 0 and mentions ARCHITECTURE-MAP.json ---
+help_output=$("$ARCH_MAP" --help 2>&1) || true
+assert_exit "--help exits 0" 0 "$ARCH_MAP" --help
+assert_contains "--help mentions ARCHITECTURE-MAP.json" "ARCHITECTURE-MAP.json" "$help_output"
+
+# --- Test: generates valid JSON from a temp project ---
+TEST_TMPDIR=$(mktemp -d)
+trap 'rm -rf "$TEST_TMPDIR"' EXIT
+
+# Create a minimal project with shell + python files
+mkdir -p "$TEST_TMPDIR/lib" "$TEST_TMPDIR/src"
+cat > "$TEST_TMPDIR/main.sh" <<'SH'
+#!/usr/bin/env bash
+source lib/helpers.sh
+echo "hello"
+SH
+cat > "$TEST_TMPDIR/lib/helpers.sh" <<'SH'
+#!/usr/bin/env bash
+helper_func() { echo "help"; }
+SH
+cat > "$TEST_TMPDIR/src/app.py" <<'PY'
+from src.utils import do_thing
+import os
+
+def main():
+    do_thing()
+PY
+cat > "$TEST_TMPDIR/src/utils.py" <<'PY'
+def do_thing():
+    pass
+PY
+
+# Run architecture-map on the temp project
+map_exit=0
+"$ARCH_MAP" --project-root "$TEST_TMPDIR" > /dev/null 2>&1 || map_exit=$?
+assert_eq "generates map exit 0" "0" "$map_exit"
+
+# Check output file exists
+output_file="$TEST_TMPDIR/docs/ARCHITECTURE-MAP.json"
+TESTS=$((TESTS + 1))
+if [[ -f "$output_file" ]]; then
+    echo "PASS: creates docs/ARCHITECTURE-MAP.json"
+else
+    echo "FAIL: docs/ARCHITECTURE-MAP.json not found"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# Check JSON is valid
+json_valid=0
+jq . "$output_file" > /dev/null 2>&1 || json_valid=1
+assert_eq "output is valid JSON" "0" "$json_valid"
+
+# Check required fields
+has_generated=$(jq 'has("generated_at")' "$output_file" 2>/dev/null)
+assert_eq "JSON has generated_at field" "true" "$has_generated"
+
+has_modules=$(jq 'has("modules")' "$output_file" 2>/dev/null)
+assert_eq "JSON has modules field" "true" "$has_modules"
+
+# Check shell source dependency detected
+source_deps=$(jq -r '[.modules[].files[]? | select(.dependencies[]? | contains("lib/helpers.sh"))] | length' "$output_file" 2>/dev/null || echo "0")
+TESTS=$((TESTS + 1))
+if [[ "$source_deps" -gt 0 ]]; then
+    echo "PASS: detects shell source dependencies"
+else
+    echo "FAIL: did not detect shell source dependency on lib/helpers.sh"
+    echo "  JSON: $(cat "$output_file")"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# --- Test: empty project produces empty modules, exits 0 ---
+EMPTY_DIR=$(mktemp -d)
+empty_exit=0
+"$ARCH_MAP" --project-root "$EMPTY_DIR" > /dev/null 2>&1 || empty_exit=$?
+assert_eq "empty project exits 0" "0" "$empty_exit"
+
+empty_modules=$(jq '.modules | length' "$EMPTY_DIR/docs/ARCHITECTURE-MAP.json" 2>/dev/null || echo "-1")
+assert_eq "empty project has 0 modules" "0" "$empty_modules"
+rm -rf "$EMPTY_DIR"
+
+report_results

package/scripts/tests/test-auto-compound.sh

@@ -0,0 +1,169 @@
+#!/usr/bin/env bash
+# Test auto-compound.sh — verifies refactored behavior using common.sh/ollama.sh
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+AC="$SCRIPT_DIR/../auto-compound.sh"
+
+FAILURES=0
+TESTS=0
+
+assert_eq() {
+    local desc="$1" expected="$2" actual="$3"
+    TESTS=$((TESTS + 1))
+    if [[ "$expected" != "$actual" ]]; then
+        echo "FAIL: $desc"
+        echo "  expected: $expected"
+        echo "  actual:   $actual"
+        FAILURES=$((FAILURES + 1))
+    else
+        echo "PASS: $desc"
+    fi
+}
+
+assert_exit() {
+    local desc="$1" expected_exit="$2"
+    shift 2
+    local actual_exit=0
+    "$@" >/dev/null 2>&1 || actual_exit=$?
+    TESTS=$((TESTS + 1))
+    if [[ "$expected_exit" != "$actual_exit" ]]; then
+        echo "FAIL: $desc"
+        echo "  expected exit: $expected_exit"
+        echo "  actual exit:   $actual_exit"
+        FAILURES=$((FAILURES + 1))
+    else
+        echo "PASS: $desc"
+    fi
+}
+
+# === Sources shared libraries ===
+
+TESTS=$((TESTS + 1))
+if grep -q 'source.*lib/common.sh' "$AC"; then
+    echo "PASS: auto-compound.sh sources lib/common.sh"
+else
+    echo "FAIL: auto-compound.sh sources lib/common.sh"
+    FAILURES=$((FAILURES + 1))
+fi
+
+TESTS=$((TESTS + 1))
+if grep -q 'source.*lib/ollama.sh' "$AC"; then
+    echo "PASS: auto-compound.sh sources lib/ollama.sh"
+else
+    echo "FAIL: auto-compound.sh sources lib/ollama.sh"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === CLI tests ===
+
+assert_exit "no args exits 1" 1 bash "$AC"
+assert_exit "--help exits 0" 0 bash "$AC" /dev/null --help
+
+# === No silent PRD discard (lesson-7) ===
+# The old line was: > /dev/null 2>&1 || true (discard all output)
+# New code should capture output and log errors
+
+TESTS=$((TESTS + 1))
+if grep -q '> /dev/null 2>&1 || true' "$AC"; then
+    echo "FAIL: PRD output still silently discarded (lesson-7 violation)"
+    FAILURES=$((FAILURES + 1))
+else
+    echo "PASS: PRD output not silently discarded"
+fi
+
+# === Uses detect_project_type in fallback ===
+
+TESTS=$((TESTS + 1))
+if grep -q 'detect_project_type' "$AC"; then
+    echo "PASS: uses detect_project_type for fallback detection"
+else
+    echo "FAIL: should use detect_project_type for fallback detection"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Bug #14: head -c 40 replaced with cut -c1-40 for UTF-8 safety ===
+
+# Verify the script uses cut -c1-40 instead of head -c 40
+TESTS=$((TESTS + 1))
+if grep -q 'cut -c1-40' "$AC"; then
+    echo "PASS: uses cut -c1-40 for UTF-8-safe character truncation"
+else
+    echo "FAIL: should use cut -c1-40 instead of head -c 40"
+    FAILURES=$((FAILURES + 1))
+fi
+
+TESTS=$((TESTS + 1))
+if grep -q 'head -c' "$AC"; then
+    echo "FAIL: still contains head -c (byte-level truncation, breaks UTF-8)"
+    FAILURES=$((FAILURES + 1))
+else
+    echo "PASS: no head -c byte truncation remaining"
+fi
+
+# Functional test: multi-byte characters are not split
+# The slug pipeline: lowercase → replace non-alnum with dash → collapse dashes → cut to 40 chars
+# CJK and emoji get replaced by dashes in the sed step, but the cut must not corrupt them
+# before sed processes them. Test the full pipeline expression.
+TESTS=$((TESTS + 1))
+SLUG_PIPELINE='tr '\''[:upper:]'\'' '\''[:lower:]'\'' | sed '\''s/[^a-z0-9]/-/g'\'' | sed '\''s/--*/-/g'\'' | cut -c1-40'
+# Input with emoji and CJK — these become dashes after sed, but cut -c handles them correctly
+INPUT="add-emoji-support-🚀🎉-for-notifications"
+RESULT=$(echo "$INPUT" | eval "$SLUG_PIPELINE")
+# Verify result is valid (no broken bytes — should be pure ASCII after sed)
+VALID=$(echo "$RESULT" | LC_ALL=C grep -c '[^ -~]' || true)
+if [[ "$VALID" == "0" ]]; then
+    echo "PASS: multi-byte input produces valid slug (no broken UTF-8)"
+else
+    echo "FAIL: multi-byte input produced invalid bytes in slug"
+    echo "  result: $RESULT"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Bug #19: ls -t replaced with find+sort for space-safe file selection ===
+
+# Verify the script uses find instead of ls -t
+TESTS=$((TESTS + 1))
+if grep -q 'find reports/' "$AC"; then
+    echo "PASS: uses find for report file selection"
+else
+    echo "FAIL: should use find instead of ls -t for report selection"
+    FAILURES=$((FAILURES + 1))
+fi
+
+TESTS=$((TESTS + 1))
+if grep 'ls -t' "$AC" | grep -qv '^#'; then
+    echo "FAIL: still contains ls -t (breaks on filenames with spaces)"
+    FAILURES=$((FAILURES + 1))
+else
+    echo "PASS: no ls -t for file selection remaining"
+fi
+
+# Functional test: filenames with spaces
+TESTS=$((TESTS + 1))
+TMPDIR_SPACES=$(mktemp -d)
+mkdir -p "$TMPDIR_SPACES/reports"
+# Create files with spaces, newest last
+echo "old report" > "$TMPDIR_SPACES/reports/old report.md"
+sleep 0.1
+echo "new report" > "$TMPDIR_SPACES/reports/my new report.md"
+# Use the same find command from auto-compound.sh
+FOUND=$(cd "$TMPDIR_SPACES" && find reports/ -name '*.md' -printf '%T@ %p\n' 2>/dev/null | sort -rn | head -1 | cut -d' ' -f2-)
+if [[ "$FOUND" == "reports/my new report.md" ]]; then
+    echo "PASS: find selects newest file even with spaces in name"
+else
+    echo "FAIL: find did not select correct file with spaces"
+    echo "  expected: reports/my new report.md"
+    echo "  actual:   $FOUND"
+    FAILURES=$((FAILURES + 1))
+fi
+rm -rf "$TMPDIR_SPACES"
+
+# === Summary ===
+echo ""
+echo "Results: $((TESTS - FAILURES))/$TESTS passed"
+if [[ $FAILURES -gt 0 ]]; then
+    echo "FAILURES: $FAILURES"
+    exit 1
+fi
+echo "ALL PASSED"

package/scripts/tests/test-batch-test.sh

@@ -0,0 +1,65 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+source "$SCRIPT_DIR/test-helpers.sh"
+
+WORK=$(mktemp -d)
+trap 'rm -rf "$WORK"' EXIT
+
+# --- Setup: create fake project directories with Makefiles ---
+mkdir -p "$WORK/projects/project-a"
+mkdir -p "$WORK/projects/project-b"
+
+# Use a unique log file for tracking which directory each project runs from
+LOG_FILE="$WORK/cwd-log"
+
+cat > "$WORK/projects/project-a/Makefile" << EOF
+test:
+	@pwd >> $LOG_FILE
+	@echo "project-a tests passed"
+EOF
+
+cat > "$WORK/projects/project-b/Makefile" << EOF
+test:
+	@pwd >> $LOG_FILE
+	@echo "project-b tests passed"
+EOF
+
+# --- Test: cd is restored between project iterations ---
+# Run batch-test across both projects
+output=$("$SCRIPT_DIR/../batch-test.sh" "$WORK/projects" 2>&1) || true
+
+# Read the logged working directories
+if [[ -f "$LOG_FILE" ]]; then
+    cwd_a=$(sed -n '1p' "$LOG_FILE")
+    cwd_b=$(sed -n '2p' "$LOG_FILE")
+
+    # Each project should run from its own directory
+    assert_contains "batch-test: project-a runs from its own dir" "project-a" "$cwd_a"
+    assert_contains "batch-test: project-b runs from its own dir" "project-b" "$cwd_b"
+else
+    # If log file doesn't exist, both projects failed to run
+    TESTS=$((TESTS + 1))
+    echo "FAIL: batch-test: cwd log file not created"
+    FAILURES=$((FAILURES + 1))
+    TESTS=$((TESTS + 1))
+    echo "FAIL: batch-test: project-b dir check (no log)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# --- Test: output includes both projects ---
+assert_contains "batch-test: output includes project-a" "project-a" "$output"
+assert_contains "batch-test: output includes project-b" "project-b" "$output"
+assert_contains "batch-test: output includes Done" "Done" "$output"
+
+# --- Test: single project target works ---
+single_output=$("$SCRIPT_DIR/../batch-test.sh" "$WORK/projects" "project-a" 2>&1) || true
+assert_contains "batch-test: single target runs" "project-a" "$single_output"
+assert_not_contains "batch-test: single target skips other" "project-b" "$single_output"
+
+# --- Test: missing project is skipped ---
+skip_output=$("$SCRIPT_DIR/../batch-test.sh" "$WORK/projects" "nonexistent" 2>&1) || true
+assert_contains "batch-test: missing project skipped" "SKIP" "$skip_output"
+
+report_results

package/scripts/tests/test-benchmark-runner.sh

@@ -0,0 +1,25 @@
+#!/usr/bin/env bash
+# Test benchmarks/runner.sh
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+REPO_ROOT="$(cd "$SCRIPT_DIR/../.." && pwd)"
+RUNNER="$REPO_ROOT/benchmarks/runner.sh"
+
+source "$SCRIPT_DIR/test-helpers.sh"
+
+# --- Test 1: list shows benchmarks ---
+output=$(bash "$RUNNER" list 2>&1)
+assert_contains "list shows benchmarks" "01-rest-endpoint" "$output"
+assert_contains "list shows all 5" "05-multi-file-feature" "$output"
+
+# --- Test 2: help works ---
+output=$(bash "$RUNNER" help 2>&1)
+assert_contains "help shows usage" "Usage:" "$output"
+
+# --- Test 3: unknown benchmark fails gracefully ---
+exit_code=0
+bash "$RUNNER" run nonexistent-benchmark >/dev/null 2>&1 || exit_code=$?
+assert_eq "unknown benchmark exits non-zero" "1" "$exit_code"
+
+report_results