agentic-devtools 0.2.65__tar.gz → 0.2.67__tar.gz

{agentic_devtools-0.2.65 → agentic_devtools-0.2.67}/.github/copilot-instructions.md

@@ -157,7 +157,7 @@ Using raw commands (e.g., `git`, raw REST API calls, `pytest`) bypasses:
 | `gh pr review --approve` + manual verification | `agdt-gh-pr-approve` |
 | `gh api .../requested_reviewers -X POST` | `agdt-gh-request-copilot-review` |
 | Manual Phase 2 PR polling loop | `agdt-gh-pr-poll-ready` |
-| `gh pr merge --squash --delete-branch` + manual verification | `agdt-gh-pr-merge` |
+| `gh pr merge --rebase --delete-branch` + manual verification | `agdt-gh-pr-merge` |
 | `gh api` loop for PR review comment replies | `agdt-gh-reply-to-review-comments` |
 
 #### Exceptions — Raw Commands Still Required
@@ -177,7 +177,7 @@ Using raw commands (e.g., `git`, raw REST API calls, `pytest`) bypasses:
 | `agdt-gh-pr-state` | Fetch structured GitHub PR state | `--pr` (int) and optionally `--repo` (owner/repo); falls back to `github.pull_request_number` and `github.repo` state keys or git remote auto-detection |
 | `agdt-gh-pr-poll-ready` | Poll PR until ready to merge or blocked | `--pr` or `github.pull_request_number`; optional `--repo`, `--poll-interval`, `--max-wait`, `--background`, `--rerun-stale-checks` |
 | `agdt-gh-pr-approve` | Approve a GitHub PR with verification and retry | `--pr` (int) or `github.pull_request_number`; optional `--repo` (owner/repo), `--body` (approval comment) |
-| `agdt-gh-pr-merge` | Merge a GitHub PR with verification and retry | `--pr` (int) or `github.pull_request_number`; optional `--repo` (owner/repo), `--strategy` (squash/merge/rebase, default squash), `--no-delete-branch`; `--repo` falls back to `github.repo` state or git remote auto-detection |
+| `agdt-gh-pr-merge` | Merge a GitHub PR with verification and retry | `--pr` (int) or `github.pull_request_number`; optional `--repo` (owner/repo), `--strategy` (squash/merge/rebase, default rebase), `--no-delete-branch`; `--repo` falls back to `github.repo` state or git remote auto-detection |
 | `agdt-gh-rerun-checks` | Re-run stale/failed GitHub Actions workflows | `--pr` (int) or `github.pull_request_number`; optional `--repo`, `--head-sha`, `--filter` (workflow name substring), `--include-cancelled` |
 | `agdt-gh-request-copilot-review` | Request Copilot review with verification | `--pr` (int) or `github.pull_request_number`; optional `--repo` |
 | `agdt-gh-reply-to-review-comments` | Post batch replies to PR review comments with verification | `--pr` (int) or `github.pull_request_number`; `--review-id` (int) or `github.copilot_review_id`; `--replies-file` (path to JSON); optional `--repo` |

{agentic_devtools-0.2.65 → agentic_devtools-0.2.67}/.github/prompts/agdt.pr-merge-manager.prompt.md

@@ -18,7 +18,7 @@ Copilot review feedback, and merges the PR once all gates are green.
 | Check CI status | `agdt-gh-pr-checks-status` | `gh pr checks` + `gh api .../check-suites` |
 | Re-run stale checks | `agdt-gh-rerun-checks` | `gh api .../actions/runs/{id}/rerun` |
 | Approve PR | `agdt-gh-pr-approve` | `gh pr review --approve` + verification |
-| Merge PR | `agdt-gh-pr-merge` | `gh pr merge --squash` + verification |
+| Merge PR | `agdt-gh-pr-merge` | `gh pr merge --rebase` + verification |
 | Run tests | `agdt-test` + `agdt-task-wait` | _Do not run `pytest` directly._ |
 | Commit & push | `agdt-git-save-work` | `git commit` + `git push` |
 | Force push | `agdt-git-force-push` | `git push --force-with-lease` |

agentic_devtools-0.2.67/.github/scripts/speckit-trigger/check-analysis-gate-cli.sh

@@ -0,0 +1,105 @@
+#!/usr/bin/env bash
+#
+# check-analysis-gate-cli.sh - Thin CLI wrapper for check-analysis-gate.sh
+#
+# Usage: check-analysis-gate-cli.sh <report_path> [--mode block|draft] [--github-actions]
+#
+# Exit codes:
+#   --mode block (default):
+#     0 = gate passed (zero unresolved CRITICALs)
+#     1 = gate failed (unresolved CRITICALs or report missing/malformed)
+#
+#   --mode draft:
+#     0 = gate passed OR unresolved CRITICALs (gate_result=fail emitted for downstream)
+#     1 = report missing/empty/malformed (hard failure in all modes)
+
+# Guard: only run main logic when executed directly (not sourced)
+if [[ "${BASH_SOURCE[0]}" == "$0" ]]; then
+    set -euo pipefail
+
+    SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+    # shellcheck source=check-analysis-gate.sh
+    source "$SCRIPT_DIR/check-analysis-gate.sh"
+
+    # Parse arguments
+    REPORT_PATH=""
+    MODE="block"
+    GITHUB_ACTIONS_FLAG="false"
+
+    while [[ $# -gt 0 ]]; do
+        case "$1" in
+            --mode)
+                if [[ $# -lt 2 || "$2" == -* ]]; then
+                    echo "Error: --mode requires a value of 'block' or 'draft'" >&2
+                    echo "Usage: check-analysis-gate-cli.sh <report_path> [--mode block|draft] [--github-actions]" >&2
+                    exit 1
+                fi
+                case "$2" in
+                    block|draft)
+                        MODE="$2"
+                        ;;
+                    *)
+                        echo "Error: Invalid value for --mode: '$2' (expected 'block' or 'draft')" >&2
+                        echo "Usage: check-analysis-gate-cli.sh <report_path> [--mode block|draft] [--github-actions]" >&2
+                        exit 1
+                        ;;
+                esac
+                shift 2
+                ;;
+            --github-actions)
+                GITHUB_ACTIONS_FLAG="true"
+                shift
+                ;;
+            -*)
+                echo "Error: Unknown flag '$1'" >&2
+                echo "Usage: check-analysis-gate-cli.sh <report_path> [--mode block|draft] [--github-actions]" >&2
+                exit 1
+                ;;
+            *)
+                if [[ -z "$REPORT_PATH" ]]; then
+                    REPORT_PATH="$1"
+                else
+                    echo "Error: Unexpected argument '$1'" >&2
+                    exit 1
+                fi
+                shift
+                ;;
+        esac
+    done
+
+    if [[ -z "$REPORT_PATH" ]]; then
+        echo "Error: report_path is required" >&2
+        echo "Usage: check-analysis-gate-cli.sh <report_path> [--mode block|draft] [--github-actions]" >&2
+        exit 1
+    fi
+
+    # Call the library function with || to capture return code under set -e
+    gate_rc=0
+    check_analysis_gate "$REPORT_PATH" "$MODE" "$GITHUB_ACTIONS_FLAG" || gate_rc=$?
+
+    # Map return codes to exit codes based on mode
+    case "$gate_rc" in
+        0)
+            # Pass — exit 0 in all modes
+            exit 0
+            ;;
+        10)
+            # Unresolved CRITICALs detected
+            if [[ "$MODE" == "draft" ]]; then
+                # Draft mode: soft pass (gate_result=fail already emitted)
+                exit 0
+            else
+                # Block mode: hard fail
+                exit 1
+            fi
+            ;;
+        20)
+            # Report missing/empty/malformed — hard fail in all modes
+            exit 1
+            ;;
+        *)
+            echo "Error: Unexpected return code $gate_rc from check_analysis_gate" >&2
+            exit 1
+            ;;
+    esac
+fi

agentic_devtools-0.2.67/.github/scripts/speckit-trigger/check-analysis-gate.sh

@@ -0,0 +1,351 @@
+#!/usr/bin/env bash
+#
+# check-analysis-gate.sh - Library script for CRITICAL analysis gate
+#
+# This is a **library** script — it defines functions only and has no
+# top-level side effects.  It is sourced by generate-spec-from-issue.sh
+# and the thin CLI wrapper check-analysis-gate-cli.sh.
+#
+# Primary interface:
+#   source "check-analysis-gate.sh"
+#   check_analysis_gate <report_path> [block|draft] [github_actions_flag]
+#
+# Function return contract:
+#   return 0  = pass — zero unresolved CRITICAL findings
+#   return 10 = soft fail — one or more unresolved CRITICAL findings
+#   return 20 = hard fail — report missing, empty, or malformed
+#
+# Caller-visible variables set on return:
+#   gate_result    = "pass" | "fail"
+#   critical_count = <integer>
+#
+# stdout: GATE_RESULT_JSON:{...} line (always)
+# stderr: Human-readable summary with banner (always)
+#
+# When github_actions_flag is "true":
+#   Writes critical_count, critical_findings (JSON), and gate_result
+#   to $GITHUB_OUTPUT.
+
+# ---------------------------------------------------------------------------
+# check_analysis_gate <report_path> [mode] [github_actions_flag]
+# ---------------------------------------------------------------------------
+check_analysis_gate() {
+    local report_path="${1:-}"
+    local mode="${2:-block}"
+    local github_actions="${3:-false}"
+
+    # Initialize caller-visible variables
+    gate_result="fail"
+    critical_count=0
+
+    # Validate mode parameter
+    if [[ "$mode" != "block" ]] && [[ "$mode" != "draft" ]]; then
+        local invalid_mode_result_json
+        printf -v invalid_mode_result_json '%s' "[{\"type\":\"invalid_mode\",\"mode\":\"$mode\",\"expected\":[\"block\",\"draft\"]}]"
+        _gate_emit_result "report_parse_error" "$report_path" "$invalid_mode_result_json" "$github_actions"
+        echo "Error: invalid mode '$mode' — expected 'block' or 'draft'" >&2
+        return 20
+    fi
+
+    # ── Missing / empty report ──────────────────────────────────────────
+    if [[ -z "$report_path" ]] || [[ ! -f "$report_path" ]]; then
+        gate_result="fail"
+        critical_count=0
+        _gate_emit_result "report_missing" "$report_path" "[]" "$github_actions"
+        echo "## ❌ SpecKit: CRITICAL Gate Failed" >&2
+        echo "" >&2
+        echo "Analysis report not found: ${report_path:-<no path>}" >&2
+        return 20
+    fi
+
+    local file_size
+    file_size=$(wc -c < "$report_path" 2>/dev/null || echo "0")
+    file_size="${file_size//[[:space:]]/}"
+    if [[ "$file_size" -eq 0 ]]; then
+        gate_result="fail"
+        critical_count=0
+        _gate_emit_result "report_missing" "$report_path" "[]" "$github_actions"
+        echo "## ❌ SpecKit: CRITICAL Gate Failed" >&2
+        echo "" >&2
+        echo "Analysis report is empty: $report_path" >&2
+        return 20
+    fi
+
+    # ── Find Findings Table header ──────────────────────────────────────
+    local header_line=""
+    local severity_col=0
+    local id_col=0
+    local summary_col=0
+    local recommendation_col=0
+    local in_table=false
+    local found_header=false
+    local past_separator=false
+
+    # We'll collect unresolved CRITICAL findings
+    local findings_json="["
+    local findings_count=0
+
+    while IFS= read -r line || [[ -n "$line" ]]; do
+        # Strip trailing carriage return (Windows line endings)
+        line="${line%$'\r'}"
+
+        # Look for Findings Table header: a pipe-delimited line containing "Severity"
+        if [[ "$found_header" == "false" ]] && [[ "$line" == "|"* ]] && [[ "$line" == *"Severity"* ]]; then
+            found_header=true
+            in_table=true
+
+            # Determine column indices dynamically
+            local col_index=0
+            local IFS_OLD="$IFS"
+            IFS='|'
+            # shellcheck disable=SC2162
+            read -r -a cols <<< "$line"
+            IFS="$IFS_OLD"
+            for cell in "${cols[@]}"; do
+                # Trim whitespace and bold/italic markers
+                local trimmed
+                trimmed="$(echo "$cell" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//' | sed 's/[*_]//g')"
+                case "$trimmed" in
+                    Severity) severity_col=$col_index ;;
+                    ID)       id_col=$col_index ;;
+                    Summary)  summary_col=$col_index ;;
+                    Recommendation) recommendation_col=$col_index ;;
+                esac
+                col_index=$((col_index + 1))
+            done
+
+            # Validate we found all required columns
+            local missing_columns=()
+            if [[ "$severity_col" -eq 0 ]]; then
+                missing_columns+=("Severity")
+            fi
+            if [[ "$id_col" -eq 0 ]]; then
+                missing_columns+=("ID")
+            fi
+            if [[ "$summary_col" -eq 0 ]]; then
+                missing_columns+=("Summary")
+            fi
+            if [[ "$recommendation_col" -eq 0 ]]; then
+                missing_columns+=("Recommendation")
+            fi
+
+            if [[ "${#missing_columns[@]}" -gt 0 ]]; then
+                gate_result="fail"
+                critical_count=0
+                _gate_emit_result "report_parse_error" "$report_path" "[]" "$github_actions"
+                echo "## ❌ SpecKit: CRITICAL Gate Failed" >&2
+                echo "" >&2
+                echo "Findings Table header found but required column(s) not detected in: $report_path" >&2
+                echo "Missing column(s): ${missing_columns[*]}" >&2
+                return 20
+            fi
+            continue
+        fi
+
+        # Skip if we haven't found the header yet
+        if [[ "$found_header" == "false" ]]; then
+            continue
+        fi
+
+        # In the table — skip separator rows (all dashes)
+        if [[ "$in_table" == "true" ]] && [[ "$past_separator" == "false" ]]; then
+            if [[ "$line" == "|"* ]] && echo "$line" | grep -qE '^\|([[:space:]]*-+[[:space:]]*\|)+[[:space:]]*$'; then
+                past_separator=true
+                continue
+            fi
+        fi
+
+        # End of table: first non-pipe line after header
+        if [[ "$in_table" == "true" ]] && [[ "$line" != "|"* ]]; then
+            break
+        fi
+
+        # Skip if still before separator
+        if [[ "$past_separator" == "false" ]]; then
+            continue
+        fi
+
+        # Parse data row
+        local IFS_OLD="$IFS"
+        IFS='|'
+        # shellcheck disable=SC2162
+        read -r -a cells <<< "$line"
+        IFS="$IFS_OLD"
+
+        local severity_raw="${cells[$severity_col]:-}"
+        local id_raw="${cells[$id_col]:-}"
+        local summary_raw="${cells[$summary_col]:-}"
+        local recommendation_raw="${cells[$recommendation_col]:-}"
+
+        # Trim whitespace
+        severity_raw="$(echo "$severity_raw" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')"
+        id_raw="$(echo "$id_raw" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')"
+        summary_raw="$(echo "$summary_raw" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')"
+        recommendation_raw="$(echo "$recommendation_raw" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')"
+
+        # Normalize severity: strip bold/italic markers (* and _)
+        local severity_normalized
+        severity_normalized="$(echo "$severity_raw" | sed 's/[*_]//g')"
+
+        # Check if this is a CRITICAL finding
+        if ! echo "$severity_normalized" | grep -qi "CRITICAL"; then
+            continue
+        fi
+
+        # Check if it's resolved: ~~CRITICAL~~.*RESOLVED (case-insensitive RESOLVED)
+        if echo "$severity_normalized" | grep -qiE '~~CRITICAL~~.*RESOLVED'; then
+            continue
+        fi
+
+        # Bare ~~CRITICAL~~ without RESOLVED = unresolved
+        # At this point we know it contains CRITICAL and is NOT resolved
+
+        findings_count=$((findings_count + 1))
+
+        # Build JSON entry (safely escape via python3)
+        local escaped_json
+        escaped_json=$(python3 -c 'import json, sys; print(json.dumps({"id": sys.argv[1], "summary": sys.argv[2], "recommendation": sys.argv[3]}))' "$id_raw" "$summary_raw" "$recommendation_raw")
+
+        if [[ "$findings_count" -gt 1 ]]; then
+            findings_json="${findings_json},"
+        fi
+        findings_json="${findings_json}${escaped_json}"
+
+    done < "$report_path"
+
+    findings_json="${findings_json}]"
+
+    # ── Malformed report: no Findings Table found ───────────────────────
+    if [[ "$found_header" == "false" ]]; then
+        gate_result="fail"
+        critical_count=0
+        _gate_emit_result "report_parse_error" "$report_path" "[]" "$github_actions"
+        echo "## ❌ SpecKit: CRITICAL Gate Failed" >&2
+        echo "" >&2
+        echo "Analysis report is malformed — no Findings Table found: $report_path" >&2
+        return 20
+    fi
+
+    # ── Evaluate gate result ────────────────────────────────────────────
+    critical_count=$findings_count
+
+    if [[ "$critical_count" -eq 0 ]]; then
+        gate_result="pass"
+        _gate_emit_result "no_critical_findings" "$report_path" "[]" "$github_actions"
+        echo "## ✅ SpecKit: CRITICAL Gate Passed" >&2
+        echo "" >&2
+        echo "Zero unresolved CRITICAL findings in: $report_path" >&2
+        return 0
+    else
+        gate_result="fail"
+        _gate_emit_result "critical_findings_detected" "$report_path" "$findings_json" "$github_actions"
+        echo "## ❌ SpecKit: CRITICAL Gate Failed" >&2
+        echo "" >&2
+        echo "Found $critical_count unresolved CRITICAL finding(s) in: $report_path" >&2
+        echo "" >&2
+
+        # List each finding
+        # Re-parse the JSON for display using the helper extractors
+        # (these avoid a jq dependency per NFR-002).
+        local idx=1
+        local display_json="$findings_json"
+        # Simple extraction: iterate through findings_json
+        while [[ "$idx" -le "$critical_count" ]]; do
+            # Extract the nth finding from the JSON array via the helper.
+            local finding
+            finding=$(echo "$display_json" | _gate_extract_finding "$idx")
+            local f_id f_summary f_recommendation
+            f_id=$(echo "$finding" | _gate_extract_field "id")
+            f_summary=$(echo "$finding" | _gate_extract_field "summary")
+            f_recommendation=$(echo "$finding" | _gate_extract_field "recommendation")
+            echo "  $idx. [$f_id] $f_summary" >&2
+            echo "     → $f_recommendation" >&2
+            echo "" >&2
+            idx=$((idx + 1))
+        done
+
+        return 10
+    fi
+}
+
+# ---------------------------------------------------------------------------
+# Internal helpers
+# ---------------------------------------------------------------------------
+
+# Emit structured gate result to stdout and optionally to GITHUB_OUTPUT
+_gate_emit_result() {
+    local reason="$1"
+    local report_path="$2"
+    local findings_json="$3"
+    local github_actions="$4"
+
+    # Escape report_path for JSON
+    local path_escaped
+    path_escaped="$(echo "$report_path" | sed 's/"/\\"/g')"
+
+    local json
+    json="{\"gate_result\":\"${gate_result}\",\"reason\":\"${reason}\",\"critical_count\":${critical_count},\"report_path\":\"${path_escaped}\"}"
+
+    echo "GATE_RESULT_JSON:${json}"
+
+    # Write to GITHUB_OUTPUT when in GitHub Actions context
+    if [[ "$github_actions" == "true" ]] && [[ -n "${GITHUB_OUTPUT:-}" ]]; then
+        echo "gate_result=${gate_result}" >> "$GITHUB_OUTPUT"
+        echo "critical_count=${critical_count}" >> "$GITHUB_OUTPUT"
+        echo "critical_findings=${findings_json}" >> "$GITHUB_OUTPUT"
+    fi
+}
+
+# Extract the Nth finding (1-based) from a JSON array string using Python
+# Usage: echo "$json_array" | _gate_extract_finding <index>
+_gate_extract_finding() {
+    local idx="$1"
+    python3 -c '
+import json
+import sys
+
+try:
+    index = int(sys.argv[1])
+except (IndexError, ValueError):
+    sys.exit(1)
+
+try:
+    data = json.load(sys.stdin)
+except json.JSONDecodeError:
+    sys.exit(1)
+
+if not isinstance(data, list):
+    sys.exit(1)
+
+if 1 <= index <= len(data):
+    sys.stdout.write(json.dumps(data[index - 1], separators=(",", ":")))
+' "$idx"
+}
+
+# Extract a field value from a JSON object string using Python
+# Usage: echo '{"id":"F-01","summary":"text"}' | _gate_extract_field "id"
+_gate_extract_field() {
+    local field="$1"
+    python3 -c '
+import json
+import sys
+
+try:
+    obj = json.load(sys.stdin)
+except json.JSONDecodeError:
+    sys.exit(1)
+
+if not isinstance(obj, dict):
+    sys.exit(1)
+
+value = obj.get(sys.argv[1], "")
+if value is None:
+    sys.stdout.write("")
+elif isinstance(value, str):
+    sys.stdout.write(value)
+elif isinstance(value, (int, float)) and not isinstance(value, bool):
+    sys.stdout.write(str(value))
+else:
+    sys.stdout.write("")
+' "$field"
+}

{agentic_devtools-0.2.65 → agentic_devtools-0.2.67}/.github/scripts/speckit-trigger/create-spec-pr.sh

@@ -41,6 +41,8 @@ BASE_BRANCH="${BASE_BRANCH:-main}"
 # Parse optional named arguments (after positional args)
 PHASE_NUMBER=""
 PHASE_NAME=""
+CREATE_DRAFT=""
+CRITICAL_FINDINGS_JSON=""
 shift 4  # consume required positional args
 # Only consume labels_json if the next arg exists and is not a named flag
 if [[ $# -gt 0 && "$1" != --* ]]; then
@@ -68,6 +70,18 @@ while [[ $# -gt 0 ]]; do
             PHASE_NAME="$2"
             shift 2
             ;;
+        --draft)
+            CREATE_DRAFT="true"
+            shift
+            ;;
+        --critical-findings-json)
+            if [[ $# -lt 2 || "$2" == --* ]]; then
+                echo "Error: --critical-findings-json requires a value" >&2
+                exit 1
+            fi
+            CRITICAL_FINDINGS_JSON="$2"
+            shift 2
+            ;;
         *)
             echo "Error: Unknown argument '$1'" >&2
             exit 1
@@ -279,20 +293,64 @@ EOF
 )
 fi
 
+# Prepend CRITICAL findings warning to PR body when creating a draft with findings
+if [[ "$CREATE_DRAFT" == "true" ]] && [[ -n "$CRITICAL_FINDINGS_JSON" ]] && [[ "$CRITICAL_FINDINGS_JSON" != "[]" ]]; then
+    # Build findings table from JSON
+    FINDINGS_WARNING="## ⚠️ CRITICAL Findings
+
+> **This PR was created as a draft** because the analysis phase detected unresolved CRITICAL findings.
+> Address these findings before marking the PR as ready for review.
+
+| ID | Summary | Recommendation |
+|---|---|---|
+"
+    # Parse JSON array using jq (available on ubuntu-latest)
+    while IFS= read -r line; do
+        FINDINGS_WARNING="${FINDINGS_WARNING}${line}
+"
+    done < <(echo "$CRITICAL_FINDINGS_JSON" | jq -r '
+        def mdcell:
+            tostring
+            | gsub("\r\n|\n|\r"; " ")
+            | gsub("\\|"; "\\\\|");
+        .[] | "| \(.id | mdcell) | \(.summary | mdcell) | \(.recommendation | mdcell) |"
+    ' 2>/dev/null || echo "| — | _Could not parse findings_ | — |")
+
+    FINDINGS_WARNING="${FINDINGS_WARNING}
+---
+
+"
+    PR_BODY="${FINDINGS_WARNING}${PR_BODY}"
+fi
+
+# Build gh pr create command arguments
+GH_CREATE_ARGS=(
+    --title "$PR_TITLE"
+    --body "$PR_BODY"
+    --base "$BASE_BRANCH"
+    --head "$BRANCH_NAME"
+)
+
+if [[ "$CREATE_DRAFT" == "true" ]]; then
+    GH_CREATE_ARGS+=(--draft)
+    echo "Creating DRAFT pull request (CRITICAL findings detected)..."
+else
+    echo "Creating pull request..."
+fi
+
 # Create the PR
-echo "Creating pull request..."
-PR_URL=$(gh pr create \
-    --title "$PR_TITLE" \
-    --body "$PR_BODY" \
-    --base "$BASE_BRANCH" \
-    --head "$BRANCH_NAME" \
-    2>&1) || {
+PR_URL=$(gh pr create "${GH_CREATE_ARGS[@]}" 2>&1) || {
     echo "Warning: Failed to create PR" >&2
     echo "Error: $PR_URL" >&2
     echo "pr_url=" >> "${GITHUB_OUTPUT:-/dev/stdout}"
     exit 0
 }
 
+# Output draft status
+if [[ "$CREATE_DRAFT" == "true" ]]; then
+    echo "is_draft=true" >> "${GITHUB_OUTPUT:-/dev/stdout}"
+fi
+
 echo "✓ Pull request created: $PR_URL"
 
 # Extract PR number from URL

agentic_devtools-0.2.67/.github/scripts/speckit-trigger/fixtures/analysis-report-dynamic-header.md

@@ -0,0 +1,23 @@
+# Spec 999 — Test Feature: Cross-Artifact Consistency & Quality Analysis
+
+---
+
+## 1. Findings Table
+
+| ID | Pass | Severity | Location(s) | Summary | Recommendation |
+|----|------|----------|-------------|---------|----------------|
+| F01 | E — Coverage Gap | **CRITICAL** | Spec §FR-001 | Missing task coverage for core requirement | Add implementation task |
+| F02 | A — Resolved | **INFO** | Plan §Phase 1 | Historical note | No action required |
+| F03 | F — Inconsistency | **HIGH** | Tasks §T005 | Dependency ordering issue | Reorder tasks |
+
+---
+
+## 2. Metrics
+
+| Metric | Value |
+|--------|-------|
+| Critical Issues Count | 1 |
+| Total Findings | 3 |
+
+---
+*Generated by Copilot SDK*

agentic_devtools-0.2.67/.github/scripts/speckit-trigger/fixtures/analysis-report-formatting-variants.md

@@ -0,0 +1,30 @@
+# Cross-Artifact Consistency & Quality Analysis Report
+
+**Feature**: Test Feature (#999)
+**Artifacts Analyzed**: spec.md, plan.md, tasks.md
+
+---
+
+## 1. Findings Table
+
+| ID | Category | Severity | Location(s) | Summary | Recommendation |
+|----|----------|----------|-------------|---------|----------------|
+| F-01 | E. Coverage Gap | CRITICAL | Spec §FR-001 | Plain CRITICAL without bold | Add task |
+| F-02 | F. Inconsistency | **CRITICAL** | Plan §Phase 2 | Bold CRITICAL | Fix inconsistency |
+| F-03 | E. Coverage Gap | ***CRITICAL*** | Spec §FR-003 | Bold+italic CRITICAL | Add coverage |
+| F-04 | A. Duplication | ~~**CRITICAL**~~ **RESOLVED** | FR-004 | Bold-inside-strikethrough resolved | No action |
+| F-05 | B. Ambiguity | **~~CRITICAL~~** RESOLVED | NFR-001 | Strikethrough-inside-bold resolved | No action |
+| F-06 | C. Underspecification | ~~CRITICAL~~ → RESOLVED | US2 | Plain strikethrough resolved | No action |
+| F-07 | F. Inconsistency | **HIGH** | Plan §Phase 5 | Non-critical finding | Align terminology |
+
+---
+
+## 2. Metrics
+
+| Metric | Value |
+|--------|-------|
+| Critical Issues Count | 3 |
+| Total Findings | 7 |
+
+---
+*Generated by Copilot SDK*

agentic_devtools-0.2.67/.github/scripts/speckit-trigger/fixtures/analysis-report-malformed-no-table.md

@@ -0,0 +1,22 @@
+# Cross-Artifact Consistency & Quality Analysis Report
+
+**Feature**: Test Feature (#999)
+**Artifacts Analyzed**: spec.md, plan.md, tasks.md
+
+---
+
+## 1. Summary
+
+This analysis found several issues but the findings table is missing due to a formatting error.
+
+The specification has some gaps that should be addressed before implementation proceeds.
+
+## 2. Metrics
+
+| Metric | Value |
+|--------|-------|
+| Total Requirements | 5 |
+| Total Tasks | 8 |
+
+---
+*Generated by Copilot SDK*