prizmkit 1.1.113 → 1.1.115

package/bundled/dev-pipeline/scripts/generate-bootstrap-prompt.py

@@ -29,7 +29,7 @@ import sys
 
 from utils import enrich_global_context, helper_replacements, load_json_file, resolve_prompt_platform, setup_logging
 from continuation import add_continuation_args, append_continuation_handoff, is_continuation
-from prompt_framework import browser_enabled_from_item, detect_dev_server_port
+from prompt_framework import browser_enabled_from_item, checkpoint_group, checkpoint_skip, checkpoint_step, detect_dev_server_port, generate_checkpoint_from_sections, merge_checkpoint_state as merge_shared_checkpoint_state
 
 
 DEFAULT_MAX_RETRIES = 3
@@ -103,12 +103,6 @@ def parse_args():
         default=None,
         help="Override pipeline mode (default: auto-detect from complexity)",
     )
-    parser.add_argument(
-        "--critic",
-        choices=["true", "false"],
-        default=None,
-        help="Override critic enablement (default: read from feature field)",
-    )
     parser.add_argument(
         "--extract-baselines",
         action="store_true",
@@ -520,12 +514,11 @@ def process_conditional_blocks(content, resume_phase):
     return content
 
 
-def process_mode_blocks(content, pipeline_mode, init_done, critic_enabled=False,
+def process_mode_blocks(content, pipeline_mode, init_done,
                         browser_interaction=False, browser_tool="auto"):
-    """Process pipeline mode, init, critic, and browser conditional blocks.
+    """Process pipeline mode, init, and browser conditional blocks.
 
     Keeps the block matching the current mode, removes the others.
-    Handles {{IF_CRITIC_ENABLED}} / {{END_IF_CRITIC_ENABLED}} blocks.
     Handles {{IF_BROWSER_INTERACTION}} / {{END_IF_BROWSER_INTERACTION}} blocks.
     Handles {{IF_BROWSER_TOOL_PLAYWRIGHT}} / {{IF_BROWSER_TOOL_OPENCLI}} /
     {{IF_BROWSER_TOOL_AUTO}} blocks (nested inside browser interaction block).
@@ -564,20 +557,6 @@ def process_mode_blocks(content, pipeline_mode, init_done, critic_enabled=False,
             "", content, flags=re.DOTALL,
         )
 
-    # Critic blocks
-    critic_open = "{{IF_CRITIC_ENABLED}}"
-    critic_close = "{{END_IF_CRITIC_ENABLED}}"
-    if critic_enabled:
-        # Keep content, remove tags
-        content = content.replace(critic_open + "\n", "")
-        content = content.replace(critic_open, "")
-        content = content.replace(critic_close + "\n", "")
-        content = content.replace(critic_close, "")
-    else:
-        # Remove entire CRITIC blocks
-        pattern = re.escape(critic_open) + r".*?" + re.escape(critic_close) + r"\n?"
-        content = re.sub(pattern, "", content, flags=re.DOTALL)
-
     # Browser interaction blocks
     browser_open = "{{IF_BROWSER_INTERACTION}}"
     browser_close = "{{END_IF_BROWSER_INTERACTION}}"
@@ -672,36 +651,41 @@ def determine_pipeline_mode(complexity):
 # prizmkit skill name.  This ensures each section has a distinct key so
 # merge_checkpoint_state() never collides.
 SECTION_TO_SKILL = {
-    "phase0-init": ("prizmkit-init", "Project Bootstrap",
-                    [".prizmkit/prizm-docs/root.prizm", ".prizmkit/config.json"]),
-    "phase0-test-baseline": None,
-    "phase-context-snapshot": ("context-snapshot", "Build Context Snapshot",
-                               [".prizmkit/specs/{slug}/context-snapshot.md"]),
-    "phase-specify-plan": ("context-snapshot-and-plan", "Specify & Plan",
-                           [".prizmkit/specs/{slug}/context-snapshot.md",
-                            ".prizmkit/specs/{slug}/plan.md"]),
-    "phase-plan": ("prizmkit-plan", "Plan & Tasks",
-                   [".prizmkit/specs/{slug}/plan.md"]),
-    "phase-critic-plan": ("critic-plan-review", "Critic: Plan Review", []),
-    "phase-implement": ("prizmkit-implement", "Implement + Test", []),
-    "phase-prizmkit-test": (
+    "phase0-init": checkpoint_step(
+        "prizmkit-init", "Project Bootstrap",
+        [".prizmkit/prizm-docs/root.prizm", ".prizmkit/config.json"],
+    ),
+    "phase0-test-baseline": checkpoint_skip(),
+    "phase-context-snapshot": checkpoint_step(
+        "context-snapshot", "Build Context Snapshot",
+        [".prizmkit/specs/{slug}/context-snapshot.md"],
+    ),
+    "phase-specify-plan": checkpoint_step(
+        "context-snapshot-and-plan", "Specify & Plan",
+        [".prizmkit/specs/{slug}/context-snapshot.md",
+         ".prizmkit/specs/{slug}/plan.md"],
+    ),
+    "phase-plan": checkpoint_step(
+        "prizmkit-plan", "Plan & Tasks",
+        [".prizmkit/specs/{slug}/plan.md"],
+    ),
+    "phase-implement": checkpoint_step("prizmkit-implement", "Implement + Test", []),
+    "phase-prizmkit-test": checkpoint_step(
         "prizmkit-test", "Scoped Feature Test Gate",
         [".prizmkit/specs/{slug}/test-report-path.txt",
          ".prizmkit/test/*/test-report.md"],
     ),
-    "phase-review": ("prizmkit-code-review", "Code Review", []),
-    "phase-browser": ("browser-verification", "Browser Verification", []),
-    "phase-commit": None,  # special: split into retrospective + committer
+    "phase-review": checkpoint_step("prizmkit-code-review", "Code Review", []),
+    "phase-browser": checkpoint_step("browser-verification", "Browser Verification", []),
+    "phase-commit": checkpoint_group([
+        ("prizmkit-retrospective", "Retrospective", []),
+        ("prizmkit-committer", "Commit", []),
+        ("completion-summary", "Completion Summary",
+         [".prizmkit/specs/{slug}/completion-summary.json"]),
+    ]),
 }
 
-# phase-commit is split into durable checkpoint steps. CLI flags are not
-# artifacts; only real files/globs belong in required_artifacts.
-_COMMIT_STEPS = [
-    ("prizmkit-retrospective", "Retrospective", []),
-    ("prizmkit-committer", "Commit", []),
-    ("completion-summary", "Completion Summary",
-     [".prizmkit/specs/{slug}/completion-summary.json"]),
-]
+_COMMIT_STEPS = []
 
 
 def _resolve_artifacts(artifact_templates, slug):
@@ -1152,132 +1136,30 @@ def _completed_step_artifacts_valid(project_root, skill_key, artifacts):
 def generate_checkpoint_definition(sections, pipeline_mode, workflow_type,
                                    item_id, item_slug, session_id,
                                    init_done=False):
-    """Derive checkpoint step definitions from the assembled sections list.
-
-    Args:
-        sections: list of (name, content) tuples from assemble_sections()
-        pipeline_mode: "lite" | "standard" | "full"
-        workflow_type: "feature-pipeline"
-        item_id: feature ID (e.g. "F-001")
-        item_slug: feature slug (e.g. "001-user-auth")
-        session_id: current session ID
-        init_done: whether project is already initialized (Phase 0 skip)
-
-    Returns:
-        dict suitable for writing as workflow-checkpoint.json
-    """
-    steps = []
-    step_counter = 1
-    prev_step_id = None
-
-    for section_name, _content in sections:
-        if section_name not in SECTION_TO_SKILL:
-            continue
-
-        mapping = SECTION_TO_SKILL[section_name]
-
-        if mapping is None:
-            # phase-commit -> split into retrospective + committer
-            for skill, name, artifacts in _COMMIT_STEPS:
-                step_id = "S{:02d}".format(step_counter)
-                steps.append({
-                    "id": step_id,
-                    "skill": skill,
-                    "name": name,
-                    "status": "pending",
-                    "required_artifacts": _resolve_artifacts(artifacts, item_slug),
-                    "depends_on": prev_step_id,
-                })
-                prev_step_id = step_id
-                step_counter += 1
-            continue
-
-        skill, name, artifacts = mapping
-        step_id = "S{:02d}".format(step_counter)
-
-        status = "pending"
-        if init_done and section_name in ("phase0-init", "phase0-test-baseline"):
-            status = "skipped"
-
-        steps.append({
-            "id": step_id,
-            "skill": skill,
-            "name": name,
-            "status": status,
-            "required_artifacts": _resolve_artifacts(artifacts, item_slug),
-            "depends_on": prev_step_id,
-        })
-
-        prev_step_id = step_id
-        step_counter += 1
-
-    return {
-        "version": 1,
-        "workflow_type": workflow_type,
-        "pipeline_mode": pipeline_mode,
-        "item_id": item_id,
-        "item_slug": item_slug,
-        "session_id": session_id,
-        "steps": steps,
-    }
+    """Derive checkpoint step definitions from explicit section contracts."""
+    return generate_checkpoint_from_sections(
+        sections,
+        SECTION_TO_SKILL,
+        pipeline_mode,
+        workflow_type,
+        item_id,
+        item_slug,
+        session_id,
+        init_done=init_done,
+    )
 
 
 def merge_checkpoint_state(existing, fresh, project_root):
-    """Merge existing checkpoint state into a freshly generated definition.
-
-    Matching is by skill_key (not step ID), since tier changes across retries
-    may shift step IDs.
-
-    Merge rules:
-    1. Only keep completed steps whose required_artifacts all exist on disk
-    2. Keep skipped steps unconditionally
-    3. Once a step is NOT completed/skipped, break the dependency chain:
-       all subsequent steps reset to pending
-    """
-    existing_status = {}
-    existing_artifacts = {}
-    fresh_artifacts = {step["skill"]: step.get("required_artifacts", [])
-                       for step in fresh.get("steps", [])}
-    for step in existing.get("steps", []):
-        existing_status[step["skill"]] = step["status"]
-        existing_artifacts[step["skill"]] = step.get("required_artifacts", [])
-
-    # Determine which completed steps have valid artifacts. Prefer the freshly
-    # generated artifact contract so newly added gates invalidate older
-    # checkpoint entries that only satisfied a weaker contract.
-    valid_completed = set()
-    for skill_key, status in existing_status.items():
-        if status == "completed":
-            artifacts = fresh_artifacts.get(
-                skill_key, existing_artifacts.get(skill_key, [])
-            )
-            if _completed_step_artifacts_valid(project_root,
-                                               skill_key,
-                                               artifacts):
-                valid_completed.add(skill_key)
-            else:
-                LOGGER.warning(
-                    "Step '%s' was completed but artifacts missing — "
-                    "resetting to pending", skill_key
-                )
-        elif status == "skipped":
-            valid_completed.add(skill_key)
-
-    # Apply to fresh steps; break chain on first non-valid step
-    chain_broken = False
-    for step in fresh["steps"]:
-        if chain_broken:
-            step["status"] = "pending"
-            continue
-
-        prev = existing_status.get(step["skill"])
-        if step["skill"] in valid_completed:
-            step["status"] = prev  # completed or skipped
-        else:
-            chain_broken = True
-            step["status"] = "pending"
-
-    return fresh
+    """Merge existing checkpoint state into a freshly generated definition."""
+    return merge_shared_checkpoint_state(
+        existing,
+        fresh,
+        project_root,
+        validator=lambda skill_key, artifacts: _completed_step_artifacts_valid(
+            project_root, skill_key, artifacts,
+        ),
+        logger=LOGGER,
+    )
 
 
 # ============================================================
@@ -1297,18 +1179,15 @@ def load_section(sections_dir, name):
         return f.read()
 
 
-ACTIVE_AGENT_PROMPTS = {
-    "critic-plan-challenge.md": "{{AGENT_PROMPT_CRITIC_PLAN_CHALLENGE}}",
-}
+ACTIVE_AGENT_PROMPTS = {}
 
 
 def load_active_agent_prompts(templates_dir):
     """Load explicit active agent prompt templates.
 
-    Retired Dev implementation and review-fix prompts are intentionally not
-    loaded; feature implementation and accepted review fixes are performed by
-    the main orchestrator. Remaining prompt placeholders support Critic and
-    read-only Reviewer handoffs.
+    Retired Dev implementation, review-fix, and plan-challenge prompts are
+    intentionally not loaded; feature implementation and accepted review fixes
+    are performed by the main orchestrator and reviewer gate.
     """
     agent_prompts_dir = os.path.join(templates_dir, "agent-prompts")
     if not os.path.isdir(agent_prompts_dir):
@@ -1341,26 +1220,24 @@ def _tier_header(pipeline_mode):
         ),
         "standard": (
             "# Dev-Pipeline Session Bootstrap — Tier 2 "
-            "(Orchestrator + Critic/Reviewer)\n",
-            "**Tier 2 — Orchestrator + Critic/Reviewer**: You handle "
-            "context + planning + implementation directly. Critic challenges "
-            "plan (if enabled), Reviewer reviews. No Dev subagent — "
-            "orchestrator implements directly (P0-6).\n",
+            "(Orchestrator + Reviewer)\n",
+            "**Tier 2 — Orchestrator + Reviewer**: You handle "
+            "context + planning + implementation directly. Reviewer reviews. "
+            "No Dev subagent — orchestrator implements directly (P0-6).\n",
         ),
         "full": (
             "# Dev-Pipeline Session Bootstrap — Tier 3 "
-            "(Orchestrator + Critic/Reviewer)\n",
-            "**Tier 3 — Orchestrator + Critic/Reviewer**: For complex "
+            "(Orchestrator + Reviewer)\n",
+            "**Tier 3 — Orchestrator + Reviewer**: For complex "
             "features, the main orchestrator handles context, planning, "
-            "and implementation directly. Critic challenges the plan "
-            "when enabled, Reviewer reviews, and no top-level Dev "
-            "implementation subagent is spawned.\n",
+            "and implementation directly. Reviewer reviews, and no top-level "
+            "Dev implementation subagent is spawned.\n",
         ),
     }
     return headers.get(pipeline_mode, headers["lite"])
 
 
-def _tier_reminders(pipeline_mode, critic_enabled=False):
+def _tier_reminders(pipeline_mode):
     """Return tier-specific reminder text."""
     common = [
         "- MANDATORY skills: `/prizmkit-retrospective`, `/prizmkit-committer` "
@@ -1384,9 +1261,8 @@ def _tier_reminders(pipeline_mode, critic_enabled=False):
         ]
     elif pipeline_mode == "standard":
         specific = [
-            "- Tier 2: orchestrator builds context+plan, Critic challenges "
-            "plan (if enabled), orchestrator implements directly (no Dev "
-            "subagent), Reviewer reviews — see P0-6",
+            "- Tier 2: orchestrator builds context+plan, implements directly "
+            "(no Dev subagent), and Reviewer reviews — see P0-6",
             "- context-snapshot.md is append-only: orchestrator writes "
             "Sections 1-4 + Implementation Log, Reviewer appends Review Notes",
             "- Gate checks enforce Implementation Log (with Gate Evidence) "
@@ -1399,8 +1275,7 @@ def _tier_reminders(pipeline_mode, critic_enabled=False):
     else:  # full
         specific = [
             "- Tier 3: full orchestration — orchestrator implements "
-            "directly (no top-level Dev subagent), Critic challenges "
-            "plan, Reviewer reviews",
+            "directly (no top-level Dev subagent), and Reviewer reviews",
             "- context-snapshot.md is append-only: orchestrator writes "
             "Sections 1-4 + Implementation Log, Reviewer appends "
             "Review Notes",
@@ -1417,8 +1292,7 @@ def _tier_reminders(pipeline_mode, critic_enabled=False):
 
 
 def assemble_sections(pipeline_mode, sections_dir, init_done, is_resume,
-                      critic_enabled, browser_enabled, retry_count=0,
-                      browser_tool="auto"):
+                      browser_enabled, retry_count=0, browser_tool="auto"):
     """Assemble prompt sections based on tier and conditions.
 
     Uses Python code for conditional logic instead of regex-based
@@ -1535,16 +1409,6 @@ def assemble_sections(pipeline_mode, sections_dir, init_done, is_resume,
                               load_section(sections_dir,
                                            "phase-plan-agent.md")))
 
-    # --- Critic: Plan Challenge (only if critic enabled) ---
-    if critic_enabled:
-        if pipeline_mode == "full":
-            sections.append(("phase-critic-plan",
-                              load_section(sections_dir,
-                                           "phase-critic-plan-full.md")))
-        else:
-            sections.append(("phase-critic-plan",
-                              load_section(sections_dir,
-                                           "phase-critic-plan.md")))
 
     # --- Implement (tier-dependent) ---
     if pipeline_mode == "lite":
@@ -1626,7 +1490,7 @@ def assemble_sections(pipeline_mode, sections_dir, init_done, is_resume,
 
     # --- Reminders ---
     sections.append(("reminders",
-                      _tier_reminders(pipeline_mode, critic_enabled)))
+                      _tier_reminders(pipeline_mode)))
 
     return sections
 
@@ -1639,10 +1503,8 @@ def render_from_sections(sections, replacements):
     """
     content = "\n".join(text for _, text in sections)
 
-    # Replace all placeholders — run twice to handle active agent prompt
-    # templates that contain their own {{PLACEHOLDER}} variables. First pass
-    # injects prompt content (for example, Critic plan challenge text); second
-    # pass replaces inner feature/session variables.
+    # Replace all placeholders — run twice to handle active prompt fragments
+    # that may contain their own {{PLACEHOLDER}} variables.
     for _pass in range(2):
         for placeholder, value in replacements.items():
             content = content.replace(placeholder, value)
@@ -1746,9 +1608,6 @@ def build_replacements(args, feature, features, global_context, script_dir):
     reviewer_subagent = os.path.join(
         agents_dir, f"prizm-dev-team-reviewer{agent_ext}",
     )
-    critic_subagent = os.path.join(
-        agents_dir, f"prizm-dev-team-critic{agent_ext}",
-    )
 
     # Verify agent files actually exist — missing files cause confusing
     # errors when the AI session tries to read them later.
@@ -1802,40 +1661,6 @@ def build_replacements(args, feature, features, global_context, script_dir):
     if effective_resume == "null" and artifacts["all_complete"]:
         effective_resume = "6"
 
-    # Determine critic enablement (priority: CLI > env > feature field > default)
-    critic_env = os.environ.get("ENABLE_CRITIC", "").lower()
-    if args.critic is not None:
-        critic_enabled = args.critic == "true"
-    elif critic_env in ("true", "1"):
-        critic_enabled = True
-    elif critic_env in ("false", "0"):
-        critic_enabled = False
-    else:
-        critic_enabled = bool(feature.get("critic", False))
-
-    # Determine critic count (from feature field, default 1)
-    # Multi-critic voting (3) must be explicitly set by the user in .prizmkit/plans/feature-list.json
-    critic_count = feature.get("critic_count", 1)
-
-    # Guard: if critic enabled but agent file missing, force disable and warn
-    if critic_enabled and not os.path.isfile(critic_subagent):
-        LOGGER.warning(
-            "Critic enabled but agent file not found: %s. "
-            "Critic phases will be SKIPPED. "
-            "Run `npx prizmkit install` to install agent definitions.",
-            critic_subagent,
-        )
-        critic_enabled = False
-
-    # Guard: if critic enabled but tier doesn't support it (lite), warn and disable
-    if critic_enabled and pipeline_mode == "lite":
-        LOGGER.warning(
-            "Critic enabled for feature %s but pipeline_mode='lite' (tier1) "
-            "does not support critic phases. Critic will be SKIPPED. "
-            "Use estimated_complexity='high' or pass --mode standard/full.",
-            args.feature_id,
-        )
-        critic_enabled = False
 
     # Browser verification defaults to a mandatory attempt. It is disabled only
     # by BROWSER_VERIFY=false or browser_interaction.enabled=false.
@@ -1888,7 +1713,6 @@ def build_replacements(args, feature, features, global_context, script_dir):
         "{{TEAM_CONFIG_PATH}}": team_config_path,
         "{{DEV_SUBAGENT_PATH}}": dev_subagent,
         "{{REVIEWER_SUBAGENT_PATH}}": reviewer_subagent,
-        "{{CRITIC_SUBAGENT_PATH}}": critic_subagent,
         "{{INIT_SCRIPT_PATH}}": init_script_path,
         "{{SESSION_STATUS_PATH}}": session_status_abs,
         "{{PROJECT_ROOT}}": project_root,
@@ -1899,8 +1723,6 @@ def build_replacements(args, feature, features, global_context, script_dir):
         ),
         "{{PIPELINE_MODE}}": pipeline_mode,
         "{{COMPLEXITY}}": complexity,
-        "{{CRITIC_ENABLED}}": "true" if critic_enabled else "false",
-        "{{CRITIC_COUNT}}": str(critic_count),
         "{{INIT_DONE}}": "true" if init_done else "false",
         "{{HAS_SPEC}}": "true" if artifacts["has_spec"] else "false",
         "{{HAS_PLAN}}": "true" if artifacts["has_plan"] else "false",
@@ -1932,11 +1754,10 @@ def render_template(template_content, replacements, resume_phase,
     # Step 1: Process fresh_start/resume conditional blocks
     content = process_conditional_blocks(template_content, resume_phase)
 
-    # Step 2: Process mode, init, critic, and browser conditional blocks
+    # Step 2: Process mode, init, and browser conditional blocks
     pipeline_mode = replacements.get("{{PIPELINE_MODE}}", "standard")
     init_done = replacements.get("{{INIT_DONE}}", "false") == "true"
-    critic_enabled = replacements.get("{{CRITIC_ENABLED}}", "false") == "true"
-    content = process_mode_blocks(content, pipeline_mode, init_done, critic_enabled,
+    content = process_mode_blocks(content, pipeline_mode, init_done,
                                   browser_enabled, browser_tool)
 
     # Step 3: Replace all {{PLACEHOLDER}} variables (two passes for nested
@@ -2012,8 +1833,6 @@ def main():
     pipeline_mode = replacements.get("{{PIPELINE_MODE}}", "lite")
     init_done = replacements.get("{{INIT_DONE}}", "false") == "true"
     is_resume = effective_resume != "null"
-    critic_enabled = replacements.get("{{CRITIC_ENABLED}}", "false") == "true"
-
     # ── Choose rendering path ──────────────────────────────────────────
     use_sections = os.path.isdir(sections_dir) and not args.template
 
@@ -2023,7 +1842,7 @@ def main():
         try:
             sections = assemble_sections(
                 pipeline_mode, sections_dir, init_done, is_resume,
-                critic_enabled, browser_enabled,
+                browser_enabled,
                 retry_count=int(args.retry_count),
                 browser_tool=browser_tool,
             )
@@ -2140,16 +1959,12 @@ def main():
     feature_model = feature.get("model", "")
     mode_agent_counts = {"lite": 1, "standard": 3, "full": 3}
     agent_count = mode_agent_counts.get(pipeline_mode, 1)
-    critic_count_val = int(replacements.get("{{CRITIC_COUNT}}", "1"))
-    if critic_enabled:
-        agent_count += critic_count_val
     output = {
         "success": True,
         "output_path": os.path.abspath(args.output),
         "model": feature_model,
         "pipeline_mode": pipeline_mode,
         "agent_count": agent_count,
-        "critic_enabled": "true" if critic_enabled else "false",
         "render_mode": "sections" if use_sections else "legacy",
         "validation_warnings": len(warnings),
         "validation_errors": len(errors),