prizmkit 1.1.139 → 1.1.141

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

@@ -40,15 +40,14 @@ def route_bugfix_repair(stage_result, repair_scope=None, repair_round=0):
     return route_pipeline_repair(stage_result, repair_scope, repair_round)
 
 
-def bugfix_checkpoint_metadata(artifact_dir, evidence=None, stage="plan",
+def bugfix_checkpoint_metadata(artifact_dir, stage="plan",
                                status="PLAN_READY", stage_result=None,
                                repair_scope=None, repair_round=0,
                                next_stage="prizmkit-implement",
                                resume_from="prizmkit-implement"):
-    """Build Bugfix L4 handoff metadata while preserving Bugfix L1 state."""
+    """Build Bugfix pipeline handoff metadata while preserving skill state."""
     return pipeline_checkpoint_metadata(
         artifact_dir,
-        evidence=evidence,
         stage=stage,
         status=status,
         stage_result=stage_result,
@@ -71,9 +70,8 @@ BUGFIX_SECTION_TO_SKILL = {
     "bugfix-phase-test": checkpoint_step(
         "prizmkit-test", "Post-Review Test Gate",
         [
-            ".prizmkit/test/evidence/*/manifest.json",
-            ".prizmkit/test/evidence/*/verdict.json",
-            ".prizmkit/test/evidence/*/validation.json",
+            ".prizmkit/bugfix/{slug}/test-report.md",
+            ".prizmkit/bugfix/{slug}/test-result.json",
         ],
     ),
     "phase-browser": checkpoint_step("browser-verification", "Browser Verification", []),

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

@@ -67,7 +67,7 @@ affected files, and root cause hypothesis.""",
         "Triage",
         """\
 Locate affected code and identify root cause:
-1. Read `.prizmkit/prizm-docs/root.prizm` then relevant L1/L2 docs for affected modules
+1. Read `.prizmkit/prizm-docs/root.prizm` then the relevant module indexes and detail docs for affected modules
 2. Read files mentioned in the bug description or error/stack trace
 3. Check `.prizmkit/prizm-docs/` TRAPS for known patterns
 4. Classify: root cause (confirmed/suspected), blast radius, fix complexity

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

@@ -40,15 +40,14 @@ def route_refactor_repair(stage_result, repair_scope=None, repair_round=0):
     return route_pipeline_repair(stage_result, repair_scope, repair_round)
 
 
-def refactor_checkpoint_metadata(artifact_dir, evidence=None, stage="plan",
+def refactor_checkpoint_metadata(artifact_dir, stage="plan",
                                  status="PLAN_READY", stage_result=None,
                                  repair_scope=None, repair_round=0,
                                  next_stage="prizmkit-implement",
                                  resume_from="prizmkit-implement"):
-    """Build Refactor L4 handoff metadata while preserving Refactor L1 state."""
+    """Build Refactor pipeline handoff metadata while preserving skill state."""
     return pipeline_checkpoint_metadata(
         artifact_dir,
-        evidence=evidence,
         stage=stage,
         status=status,
         stage_result=stage_result,
@@ -76,9 +75,8 @@ REFACTOR_SECTION_TO_SKILL = {
     "refactor-phase-test": checkpoint_step(
         "prizmkit-test", "Post-Review Test Gate",
         [
-            ".prizmkit/test/evidence/*/manifest.json",
-            ".prizmkit/test/evidence/*/verdict.json",
-            ".prizmkit/test/evidence/*/validation.json",
+            ".prizmkit/refactor/{slug}/test-report.md",
+            ".prizmkit/refactor/{slug}/test-result.json",
         ],
     ),
     "phase-browser": checkpoint_step("browser-verification", "Browser Verification", []),

package/bundled/dev-pipeline/scripts/prompt_framework.py

@@ -8,20 +8,20 @@ so feature, bugfix, and refactor prompts do not drift.
 import glob
 import os
 import re
+import sys
+
+_PIPELINE_ROOT = os.path.abspath(os.path.join(os.path.dirname(__file__), ".."))
+if _PIPELINE_ROOT not in sys.path:
+    sys.path.insert(0, _PIPELINE_ROOT)
+
+from prizmkit_runtime.test_result import valid_test_artifact_pair  # noqa: E402
 
 
 MAX_OUTER_REPAIR_ROUNDS = 3
-REPAIR_SCOPES_REQUIRING_REVIEW = {
-    "production",
-    "runtime",
-    "schema",
-    "dependency",
-    "public-interface",
-}
 
 
 def route_pipeline_repair(stage_result, repair_scope=None, repair_round=0):
-    """Map review/test outcomes to bounded, safe L4 continuation routes."""
+    """Map review/test outcomes to bounded, safe pipeline continuation routes."""
     try:
         current_round = int(repair_round)
     except (TypeError, ValueError):
@@ -46,19 +46,28 @@ def route_pipeline_repair(stage_result, repair_scope=None, repair_round=0):
         result.update(
             next_stage="prizmkit-retrospective",
             resume_from="prizmkit-retrospective",
-            reason="authoritative test evidence passed",
+            reason="project-native tests passed",
         )
         return result
     if stage_result == "TEST_BLOCKED":
         result.update(
             blocked=True,
             resumable=False,
-            reason="bounded automatic recovery cannot safely edit production",
+            reason="testing could not establish a safe verdict",
             blocked_reason="test_blocked",
             terminal_status="WORKFLOW_BLOCKED",
         )
         return result
-    if stage_result not in {"REVIEW_NEEDS_FIXES", "TEST_FAIL"}:
+    if stage_result == "TEST_NEEDS_FIXES":
+        result.update(
+            blocked=True,
+            resumable=False,
+            reason="prizmkit-test stopped with a known remaining correction",
+            blocked_reason="test_needs_fixes",
+            terminal_status="WORKFLOW_BLOCKED",
+        )
+        return result
+    if stage_result != "REVIEW_NEEDS_FIXES":
         result.update(
             blocked=True,
             reason="unknown stage result is unsafe",
@@ -76,55 +85,22 @@ def route_pipeline_repair(stage_result, repair_scope=None, repair_round=0):
         return result
 
     result["repair_round"] = current_round + 1
-    if stage_result == "REVIEW_NEEDS_FIXES":
-        result.update(
-            next_stage="prizmkit-implement",
-            resume_from="prizmkit-implement",
-            route=["prizmkit-implement", "prizmkit-code-review", "prizmkit-test"],
-            reason="review findings require implementation and fresh review",
-        )
-        return result
-    if repair_scope == "test-infrastructure":
-        result.update(
-            next_stage="prizmkit-implement",
-            resume_from="prizmkit-implement",
-            route=["prizmkit-implement", "prizmkit-test"],
-            reason="test infrastructure failure must not edit production speculatively",
-        )
-        return result
-    if repair_scope in REPAIR_SCOPES_REQUIRING_REVIEW:
-        result.update(
-            next_stage="prizmkit-implement",
-            resume_from="prizmkit-implement",
-            route=["prizmkit-implement", "prizmkit-code-review", "prizmkit-test"],
-            reason="production-facing failure requires review before retesting",
-        )
-        return result
     result.update(
-        blocked=True,
-        reason="unknown repair scope has no safe automatic route",
-        blocked_reason="unsafe_repair_scope",
-        terminal_status="WORKFLOW_BLOCKED",
+        next_stage="prizmkit-implement",
+        resume_from="prizmkit-implement",
+        route=["prizmkit-implement", "prizmkit-code-review", "prizmkit-test"],
+        reason="review findings require implementation and fresh review",
     )
     return result
 
 
-def pipeline_checkpoint_metadata(artifact_dir, evidence=None, stage="plan",
+def pipeline_checkpoint_metadata(artifact_dir, stage="plan",
                                  status="PLAN_READY", stage_result=None,
                                  repair_scope=None, repair_round=0,
                                  next_stage="prizmkit-implement",
                                  resume_from="prizmkit-implement",
                                  blocked_reason=None, terminal_status=None):
-    """Build family-neutral semantic L4 metadata without replacing L1 state."""
-    evidence = evidence or {}
-    if evidence:
-        evidence_paths = {
-            "manifest": evidence.get("manifest_path"),
-            "verdict": evidence.get("verdict_path"),
-            "validation": evidence.get("validation_path"),
-        }
-    else:
-        evidence_paths = {"manifest": None, "verdict": None, "validation": None}
+    """Build family-neutral pipeline metadata without replacing skill state."""
     current_stage = None if terminal_status else stage
     semantic_next_stage = None if terminal_status else next_stage
     return {
@@ -137,7 +113,6 @@ def pipeline_checkpoint_metadata(artifact_dir, evidence=None, stage="plan",
         "stage_result": stage_result,
         "repair_scope": repair_scope,
         "repair_round": repair_round,
-        "authoritative_evidence_paths": evidence_paths,
         "next_stage": semantic_next_stage,
         "resume_from": resume_from,
         "blocked_reason": blocked_reason,
@@ -248,6 +223,29 @@ def completed_step_artifacts_valid(project_root, artifacts):
     return all(artifact_exists(project_root, artifact) for artifact in artifacts)
 
 
+def completed_test_artifacts_valid(project_root, artifacts):
+    """Validate a sibling test-report.md and TEST_PASS test-result.json pair."""
+    reports = [item for item in artifacts if item.endswith("/test-report.md")]
+    results = [item for item in artifacts if item.endswith("/test-result.json")]
+    if len(reports) != 1 or len(results) != 1:
+        return False
+    report_path = os.path.realpath(os.path.join(project_root, reports[0]))
+    result_path = os.path.realpath(os.path.join(project_root, results[0]))
+    root_path = os.path.realpath(project_root)
+    try:
+        if os.path.commonpath([root_path, report_path]) != root_path:
+            return False
+        if os.path.commonpath([root_path, result_path]) != root_path:
+            return False
+    except ValueError:
+        return False
+    return valid_test_artifact_pair(
+        report_path,
+        result_path,
+        required_result="TEST_PASS",
+    )
+
+
 def checkpoint_skip():
     """Return an explicit no-checkpoint-step contract for a prompt section."""
     return {"kind": "skip"}
@@ -360,7 +358,13 @@ def generate_checkpoint_from_sections(
 
 def merge_checkpoint_state(existing, fresh, project_root, validator=None, logger=None):
     """Merge existing checkpoint state into a freshly generated definition."""
-    validator = validator or (lambda _skill, artifacts: completed_step_artifacts_valid(project_root, artifacts))
+    validator = validator or (
+        lambda skill, artifacts: (
+            completed_test_artifacts_valid(project_root, artifacts)
+            if skill == "prizmkit-test"
+            else completed_step_artifacts_valid(project_root, artifacts)
+        )
+    )
     existing_by_id = {}
     existing_by_skill = {}
     for step in existing.get("steps", []):
@@ -402,9 +406,6 @@ def merge_checkpoint_state(existing, fresh, project_root, validator=None, logger
                 step["status"] = "completed"
                 if stage_result:
                     step["stage_result"] = stage_result
-                evidence = previous.get("authoritative_evidence_paths")
-                if isinstance(evidence, dict):
-                    step["authoritative_evidence_paths"] = evidence
                 continue
             if logger is not None:
                 logger.warning(

package/bundled/dev-pipeline/scripts/update-bug-status.py

@@ -89,7 +89,7 @@ def parse_args():
     parser.add_argument("--continuation-summary-path", default=None, help="Optional continuation summary artifact path.")
     parser.add_argument("--no-progress-count", type=int, default=None, help="Consecutive context-overflow sessions without objective progress.")
     parser.add_argument("--progress-fingerprint", default=None, help="JSON progress fingerprint captured after the session.")
-    parser.add_argument("--checkpoint-state", default=None, help="JSON semantic L4 checkpoint snapshot captured after the session.")
+    parser.add_argument("--checkpoint-state", default=None, help="JSON semantic pipeline checkpoint snapshot captured after the session.")
     return parser.parse_args()
 
 
@@ -261,7 +261,6 @@ SEMANTIC_CHECKPOINT_FIELDS = (
     "stage_result",
     "repair_scope",
     "repair_round",
-    "authoritative_evidence_paths",
     "blocked_reason",
     "terminal_status",
 )

package/bundled/dev-pipeline/scripts/update-checkpoint.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python3
-"""Atomically update one semantic L4 workflow checkpoint."""
+"""Atomically update one semantic pipeline workflow checkpoint."""
 
 import argparse
 import json
@@ -120,13 +120,11 @@ def _mandatory_predecessor_verdict_error(steps, index, workflow_type):
     return None
 
 def _normalize_semantic_update(semantic_update):
-    values = dict(semantic_update or {})
-    evidence = values.pop("authoritative_evidence_paths", None)
-    if evidence is not None:
-        evidence = {key: value for key, value in evidence.items() if value}
-        if evidence:
-            values["authoritative_evidence_paths"] = evidence
-    return {key: value for key, value in values.items() if value is not None}
+    return {
+        key: value
+        for key, value in dict(semantic_update or {}).items()
+        if value is not None
+    }
 
 
 def _validate_semantic_update(step, semantic_update):
@@ -145,24 +143,25 @@ def _validate_semantic_update(step, semantic_update):
     terminal_status = semantic_update.get("terminal_status")
     if terminal_status is not None and terminal_status not in VALID_TERMINAL_STATUSES:
         return "Invalid terminal_status '{}'".format(terminal_status)
-    evidence = semantic_update.get("authoritative_evidence_paths")
-    if evidence is not None and set(evidence) != {"manifest", "verdict", "validation"}:
-        return "Authoritative test evidence requires manifest, verdict, and validation paths"
     if terminal_status == "WORKFLOW_BLOCKED" and not semantic_update.get("blocked_reason"):
         return "WORKFLOW_BLOCKED requires blocked_reason"
     return None
 
 
 def _apply_repair_policy(l4_state, stage_result, repair_scope):
-    if stage_result not in {"REVIEW_NEEDS_FIXES", "TEST_FAIL", "TEST_BLOCKED"}:
+    if stage_result not in {"REVIEW_NEEDS_FIXES", "TEST_NEEDS_FIXES", "TEST_BLOCKED"}:
         return
     current_round = l4_state.get("repair_round", 0)
     if not isinstance(current_round, int) or isinstance(current_round, bool):
         current_round = MAX_OUTER_REPAIR_ROUNDS
-    if stage_result == "TEST_BLOCKED":
+    if stage_result in {"TEST_NEEDS_FIXES", "TEST_BLOCKED"}:
         l4_state.update(
             terminal_status="WORKFLOW_BLOCKED",
-            blocked_reason="test_blocked",
+            blocked_reason=(
+                "test_needs_fixes"
+                if stage_result == "TEST_NEEDS_FIXES"
+                else "test_blocked"
+            ),
             current_stage=None,
             next_stage=None,
         )
@@ -202,9 +201,6 @@ def _sync_cursor(data, l4_state):
     l4_state["current_stage"] = active.get("skill") if active else None
     l4_state["next_stage"] = active.get("skill") if active else None
 
-
-
-
 def _checkpoint_project_root(checkpoint_path):
     """Resolve the project root from a conventional .prizmkit checkpoint path."""
     path = Path(os.path.abspath(checkpoint_path))
@@ -213,42 +209,6 @@ def _checkpoint_project_root(checkpoint_path):
             return str(parent.parent)
     return str(path.parent)
 
-def _discover_test_evidence_paths(checkpoint_path, data):
-    """Discover the latest finalized test package from the artifact handoff pointer."""
-    project_root = Path(_checkpoint_project_root(checkpoint_path))
-    candidates = []
-    artifact_dir = data.get("artifact_dir")
-    if isinstance(artifact_dir, str) and artifact_dir:
-        candidates.append(project_root / artifact_dir)
-    for state_key in ("feature_state", "bugfix_state", "refactor_state"):
-        state = data.get(state_key)
-        if isinstance(state, dict) and isinstance(state.get("artifact_dir"), str):
-            candidates.append(project_root / state["artifact_dir"])
-    item_slug = data.get("item_slug")
-    if isinstance(item_slug, str) and item_slug:
-        candidates.append(project_root / ".prizmkit" / "specs" / item_slug)
-        candidates.append(project_root / ".prizmkit" / "bugfix" / item_slug)
-        candidates.append(project_root / ".prizmkit" / "refactor" / item_slug)
-
-    for artifact in candidates:
-        pointer = artifact / "test-report-path.txt"
-        try:
-            report = Path(pointer.read_text(encoding="utf-8").strip()).resolve()
-        except (OSError, ValueError):
-            continue
-        evidence_dir = report.parent
-        paths = {
-            "manifest": str(evidence_dir / "manifest.json"),
-            "verdict": str(evidence_dir / "verdict.json"),
-            "validation": str(evidence_dir / "validation.json"),
-        }
-        if all(Path(value).is_file() for value in paths.values()):
-            return {
-                key: str(Path(value).relative_to(project_root))
-                for key, value in paths.items()
-            }
-    return None
-
 
 def _finalize_semantics(checkpoint_path, data, state_key):
     """Derive a deterministic terminal status after an atomic mutation."""
@@ -286,7 +246,7 @@ def update_checkpoint(
     note=None,
     semantic_update=None,
 ):
-    """Update one step and its cohesive semantic L4 state atomically."""
+    """Update one step and its cohesive semantic pipeline state atomically."""
     if new_status not in VALID_STATUSES:
         return {
             "ok": False,
@@ -323,25 +283,11 @@ def update_checkpoint(
             return {"ok": False, "error": verdict_error}
 
     semantic_update = _normalize_semantic_update(semantic_update)
-    if (
-        semantic_update.get("stage_result") == "TEST_BLOCKED"
-        and not semantic_update.get("authoritative_evidence_paths")
-    ):
-        discovered = _discover_test_evidence_paths(checkpoint_path, data)
-        if discovered:
-            semantic_update["authoritative_evidence_paths"] = discovered
     semantic_error = _validate_semantic_update(step, semantic_update)
     if semantic_error:
         return {"ok": False, "error": semantic_error}
 
     stage_result = semantic_update.get("stage_result")
-    if step.get("skill") == "prizmkit-test" and stage_result in TEST_RESULTS:
-        evidence = semantic_update.get("authoritative_evidence_paths") or step.get("authoritative_evidence_paths")
-        if not isinstance(evidence, dict) or set(evidence) != {"manifest", "verdict", "validation"}:
-            return {
-                "ok": False,
-                "error": "Every terminal test verdict requires manifest, verdict, and validation evidence paths",
-            }
     if new_status == "completed" and step.get("skill") == "prizmkit-code-review":
         stage_result = stage_result or step.get("stage_result")
         if stage_result != "REVIEW_PASS":
@@ -350,9 +296,6 @@ def update_checkpoint(
         stage_result = stage_result or step.get("stage_result")
         if stage_result != "TEST_PASS":
             return {"ok": False, "error": "Test completion requires TEST_PASS"}
-        evidence = semantic_update.get("authoritative_evidence_paths") or step.get("authoritative_evidence_paths")
-        if not isinstance(evidence, dict) or set(evidence) != {"manifest", "verdict", "validation"}:
-            return {"ok": False, "error": "Test completion requires all authoritative evidence paths"}
 
     old_status = step.get("status", "unknown")
     step["status"] = new_status
@@ -360,20 +303,6 @@ def update_checkpoint(
         step["note"] = note
     if stage_result:
         step["stage_result"] = stage_result
-    if semantic_update.get("authoritative_evidence_paths"):
-        step["authoritative_evidence_paths"] = semantic_update["authoritative_evidence_paths"]
-    if new_status == "completed" and step.get("skill") == "prizmkit-test":
-        candidate = validate_checkpoint_data(
-            data,
-            path=checkpoint_path,
-            exists=True,
-            project_root=_checkpoint_project_root(checkpoint_path),
-        )
-        if not candidate.valid or not candidate.semantic.evidence_valid:
-            return {
-                "ok": False,
-                "error": "Test completion requires validator-attested authoritative evidence",
-            }
 
     workflow_type = str(data.get("workflow_type") or "")
     state_key = semantic_state_key(workflow_type)
@@ -384,7 +313,6 @@ def update_checkpoint(
         "stage_result",
         "repair_scope",
         "repair_round",
-        "authoritative_evidence_paths",
         "blocked_reason",
         "terminal_status",
     ):
@@ -394,8 +322,6 @@ def update_checkpoint(
     l4_state["orchestrator"] = "prizmkit-l4"
     if stage_result:
         l4_state["stage_result"] = stage_result
-    if semantic_update.get("authoritative_evidence_paths"):
-        l4_state["authoritative_evidence_paths"] = semantic_update["authoritative_evidence_paths"]
     _apply_repair_policy(l4_state, stage_result, semantic_update.get("repair_scope") or l4_state.get("repair_scope"))
     _sync_cursor(data, l4_state)
     _finalize_semantics(checkpoint_path, data, state_key)
@@ -439,9 +365,6 @@ def main():
     parser.add_argument("--stage-result", choices=sorted(REVIEW_RESULTS | TEST_RESULTS), default=None)
     parser.add_argument("--repair-scope", choices=sorted(VALID_REPAIR_SCOPES), default=None)
     parser.add_argument("--repair-round", type=int, default=None)
-    parser.add_argument("--manifest-path", default=None)
-    parser.add_argument("--verdict-path", default=None)
-    parser.add_argument("--validation-path", default=None)
     parser.add_argument("--blocked-reason", default=None)
     parser.add_argument("--terminal-status", choices=sorted(VALID_TERMINAL_STATUSES), default=None)
 
@@ -450,11 +373,6 @@ def main():
         "stage_result": args.stage_result,
         "repair_scope": args.repair_scope,
         "repair_round": args.repair_round,
-        "authoritative_evidence_paths": {
-            "manifest": args.manifest_path,
-            "verdict": args.verdict_path,
-            "validation": args.validation_path,
-        },
         "blocked_reason": args.blocked_reason,
         "terminal_status": args.terminal_status,
     }

package/bundled/dev-pipeline/scripts/update-feature-status.py

@@ -156,7 +156,7 @@ def parse_args():
     parser.add_argument(
         "--checkpoint-state",
         default=None,
-        help="JSON semantic L4 checkpoint snapshot captured after the session.",
+        help="JSON semantic pipeline checkpoint snapshot captured after the session.",
     )
     return parser.parse_args()
 
@@ -370,7 +370,6 @@ SEMANTIC_CHECKPOINT_FIELDS = (
     "stage_result",
     "repair_scope",
     "repair_round",
-    "authoritative_evidence_paths",
     "blocked_reason",
     "terminal_status",
 )

package/bundled/dev-pipeline/scripts/update-refactor-status.py

@@ -101,7 +101,7 @@ def parse_args():
     parser.add_argument("--continuation-summary-path", default=None, help="Optional continuation summary artifact path.")
     parser.add_argument("--no-progress-count", type=int, default=None, help="Consecutive context-overflow sessions without objective progress.")
     parser.add_argument("--progress-fingerprint", default=None, help="JSON progress fingerprint captured after the session.")
-    parser.add_argument("--checkpoint-state", default=None, help="JSON semantic L4 checkpoint snapshot captured after the session.")
+    parser.add_argument("--checkpoint-state", default=None, help="JSON semantic pipeline checkpoint snapshot captured after the session.")
     return parser.parse_args()
 
 
@@ -260,7 +260,6 @@ SEMANTIC_CHECKPOINT_FIELDS = (
     "stage_result",
     "repair_scope",
     "repair_round",
-    "authoritative_evidence_paths",
     "blocked_reason",
     "terminal_status",
 )

package/bundled/dev-pipeline/templates/bootstrap-prompt.md

@@ -24,7 +24,7 @@ Infer what needs to be done from the feature context above and follow the active
 
 ### Execution Steps
 
-1. **Read context**: Read `.prizmkit/prizm-docs/root.prizm` and relevant L1/L2 docs to understand the codebase.
+1. **Read context**: Read `.prizmkit/prizm-docs/root.prizm` and the relevant module indexes and detail docs to understand the codebase.
 
 2. **Plan**: Run `/prizmkit-plan` with `artifact_dir=.prizmkit/specs/{{FEATURE_SLUG}}/` to produce `spec.md` and `plan.md`.
 
@@ -32,7 +32,7 @@ Infer what needs to be done from the feature context above and follow the active
 
 4. **Review**: Run `/prizmkit-code-review` with `artifact_dir=.prizmkit/specs/{{FEATURE_SLUG}}/` and follow the skill's current review contract for review execution, repairs, verification, evidence, and handoff. Require the last `## Final Result`. `REVIEW_PASS` proceeds to test; `REVIEW_NEEDS_FIXES` increments the shared outer repair round and routes implement → code-review → test. Do not invoke another review entry point or add a second review path outside the skill contract.
 
-5. **Test**: After review passes, run `/prizmkit-test scope=this-change artifact_dir=.prizmkit/specs/{{FEATURE_SLUG}}/`. Treat the Markdown report only as a derived view. Consume `.prizmkit/test/evidence/<evidence-id>/manifest.json`, `verdict.json`, and `validation.json`; verify identity/scope, every manifest hash, verdict/validator agreement, and current target hashes before accepting `TEST_PASS`. `TEST_FAIL/test-infrastructure` routes implement → test; production/runtime/schema/dependency/public-interface routes implement → code-review → test; unknown unsafe scope blocks. `TEST_BLOCKED` allows bounded environment recovery or a test-stage resume but never speculative production edits. Stop outer repairs after three rounds; preserve evidence and truthful recovery metadata.
+5. **Test**: After review passes, run `/prizmkit-test scope=this-change artifact_dir=.prizmkit/specs/{{FEATURE_SLUG}}/`. Consume the sibling `test-report.md` and `test-result.json`; they must agree. `TEST_PASS` proceeds to retrospective. `TEST_NEEDS_FIXES` or `TEST_BLOCKED` preserves its known correction or blocker and stops truthfully. Do not require an evidence package, reinterpret a testing result as a runtime crash, or recreate the skill's bounded review and repair loops.
 
 6. **Retrospective**: Run `/prizmkit-retrospective` to sync `.prizmkit/prizm-docs/` with code changes.
 
@@ -46,5 +46,5 @@ Infer what needs to be done from the feature context above and follow the active
 - If a step fails after 3 attempts, write a status report and stop.
 - **Never re-read unmodified files**: once you read a file into context, do NOT read it again unless it was modified by an Edit/Write after your last read — use `grep -n`/`sed -n` for specific lines. Before issuing a Read, ask: "Have I read this file before, and has it changed since?" If no — do NOT read it. Follow the active skill and runtime contracts when an execution unit is created; do not infer its permissions, continuation, or retry behavior from this fallback prompt.
 - **Read offset safety**: if Read returns "shorter than offset"/"empty" — run `wc -l` then `sed -n`, do NOT retry same offset.
-- **Hard Round Cap**: after 6 total test-fix rounds, STOP trial-and-error. Switch to read-once-then-rewrite. Do NOT exceed 7 rounds.
+- Follow each skill's own bounded loops; do not add a second cross-skill test-fix loop in this fallback prompt.
 - **Layered test**: debug with single-file tests, only run full suite as gate when single-file is green.

package/bundled/dev-pipeline/templates/bootstrap-tier3.md

@@ -128,7 +128,7 @@ If `EXISTING_CODE` is non-empty: your spec/plan/tasks must reflect this existing
 
 **Step A — Build Context Snapshot** (skip if `context-snapshot.md` already exists):
 
-1. Read `.prizmkit/prizm-docs/root.prizm` and relevant L1/L2 prizm docs
+1. Read `.prizmkit/prizm-docs/root.prizm` and the relevant module indexes and detail docs
 2. Detect source code directories from KEY_FILES and STRUCTURE sections in `root.prizm`; if root docs are unavailable, use the helper to list likely source files:
    ```bash
    {{RUNTIME_HELPER_CMD}} artifact source-files . --json
@@ -138,7 +138,7 @@ If `EXISTING_CODE` is non-empty: your spec/plan/tasks must reflect this existing
 4. Write `.prizmkit/specs/{{FEATURE_SLUG}}/context-snapshot.md`:
    - **Section 1 — Feature Brief**: feature description (2-3 lines) + acceptance criteria (copy from above)
    - **Section 2 — Verification Gates**: copy the gate checklist from the Task Contract above
-   - **Section 3 — Key TRAPS & RULES**: extract only relevant TRAPS/RULES from prizm-docs that apply to files in this scope. Do NOT copy full L0/L1/L2 content.
+   - **Section 3 — Key TRAPS & RULES**: extract only relevant TRAPS/RULES from the loaded Prizm docs that apply to files in this scope. Do NOT copy complete documentation files.
    - **Section 4 — File Manifest**: For each file relevant to this feature, list: file path, why it's needed (modify/reference/test), key interface signatures (function names + params + return types). Do NOT include full file content — agents read files on-demand. Format:
      ### Files to Modify
      | File | Why Needed | Key Interfaces |
@@ -151,7 +151,7 @@ If `EXISTING_CODE` is non-empty: your spec/plan/tasks must reflect this existing
      | `<source-dir>/security/permission-guard.js` | Permission check integration | `checkCommandPermission(userId, cmd)` |
 
      ### Known TRAPS (from .prizmkit/prizm-docs/)
-     - <trap entries extracted from L1/L2 docs>
+     - <trap entries extracted from the loaded Prizm docs>
 4. Confirm: `{{RUNTIME_HELPER_CMD}} artifact exists .prizmkit/specs/{{FEATURE_SLUG}}/context-snapshot.md`
 
 **After Step A**: Use context-snapshot.md Section 4 File Manifest to guide targeted file reads. Do NOT scan directories or read unrelated files.
@@ -252,26 +252,15 @@ Only `REVIEW_PASS` can proceed to the structured Feature test gate. `REVIEW_INCO
 
 ### Scoped Feature Test Gate — PrizmKit Test
 
-Create the current-run marker, then invoke the Feature gate:
-
-```bash
-{{RUNTIME_HELPER_CMD}} artifact ensure-dir .prizmkit/specs/{{FEATURE_SLUG}}
-{{RUNTIME_HELPER_CMD}} artifact touch .prizmkit/specs/{{FEATURE_SLUG}}/.prizmkit-test-started
-```
+Invoke the Feature test gate:
 
 ```text
 /prizmkit-test scope=this-change artifact_dir=.prizmkit/specs/{{FEATURE_SLUG}}/
 ```
 
-The Markdown `test-report.md` is a derived view only. Locate its package through `.prizmkit/specs/{{FEATURE_SLUG}}/test-report-path.txt` and consume authoritative `.prizmkit/test/evidence/<evidence-id>/manifest.json`, `verdict.json`, and `validation.json`. Before accepting `TEST_PASS`, verify evidence identity/scope, every manifest file hash, final-verdict agreement, validator result and payload hash, current inventory target hashes, and the compatibility value `legacy-test-report-interface-not-supported`. Never accept a Markdown-only PASS.
-
-Repair routes:
-- `TEST_FAIL` with `repair_scope=test-infrastructure` → implement → test, without speculative production edits.
-- `TEST_FAIL` with production/runtime/schema/dependency/public-interface scope → implement → code-review → test.
-- Unknown unsafe scope → `WORKFLOW_BLOCKED`; do not guess a production edit.
-- `TEST_BLOCKED` → preserve evidence/artifacts, perform bounded environment recovery, then resume from test or finish terminal `WORKFLOW_BLOCKED`; never edit production speculatively.
+Consume `.prizmkit/specs/{{FEATURE_SLUG}}/test-report.md` and `test-result.json`. They must agree. `TEST_PASS` completes the gate; `TEST_NEEDS_FIXES` preserves a known correction or delta-review requirement and stops truthfully; `TEST_BLOCKED` preserves the blocker and becomes `WORKFLOW_BLOCKED`. Do not require an evidence package, manifest, hashes, attestation, or test-internal checkpoint, and do not reinterpret a testing result as an AI CLI crash.
 
-Record `stage`, `status`, `stage_result`, `repair_scope`, `repair_round`, structured evidence paths, `next_stage`, and `resume_from` in truthful recovery metadata. Keep L1 workflow state separate from L4 `workflow-checkpoint.json`. The outer repair round stops after three and is independent of the internal review loop.
+Record only caller-owned `stage`, `status`, `stage_result`, `repair_scope`, `repair_round`, `next_stage`, and `resume_from` in recovery metadata. Keep testing reports/results separate from `workflow-checkpoint.json`. The skill owns its internal ten-round Main-Agent review, five-response optional independent review, and three-round execution-failure repair budgets.
 
 {{IF_BROWSER_INTERACTION}}
 ### Phase 5.5: Browser Verification — MANDATORY
@@ -448,7 +437,7 @@ git log --oneline
 **6b.** Run `/prizmkit-retrospective` (**before commit**, maintains `.prizmkit/prizm-docs/` architecture index):
 - **Structural sync**: update KEY_FILES/INTERFACES/DEPENDENCIES/file counts for changed modules
 - **Architecture knowledge** (feature sessions only): extract TRAPS, RULES, DECISIONS from completed work into `.prizmkit/prizm-docs/`
-- **L2 coverage check**: For any module/sub-module with source files created or significantly modified in this session but no L2 `.prizm` doc — evaluate whether L2 is warranted and create if so. The current session has the best context for accurate KEY_FILES, TRAPS, and DECISIONS.
+- **Detail-doc coverage check**: For any module/sub-module with source files created or significantly modified in this session but no detail `.prizm` doc — evaluate whether one is warranted and create it when needed. The current session has the best context for accurate KEY_FILES, TRAPS, and DECISIONS.
 - Stage doc changes: `git add .prizmkit/prizm-docs/`
 ⚠️ Do NOT commit here. Only stage.