okstra 0.49.0 → 0.51.0

package/runtime/templates/reports/i18n/en.json

@@ -15,8 +15,8 @@
     "declinedFixRecommendations": "- None.",
     "discrepancy": "- None.",
     "lingeringRisks": "- No tracked lingering risks.",
-    "noClarification": "- No additional information requested. The Section 2 verdict stands as-is.",
-    "noFollowUp": "- No follow-up tasks. The next phase for this run is in §6 (Recommended Next Steps)."
+    "noClarification": "- No additional information requested. The Section 7 verdict stands as-is.",
+    "noFollowUp": "- No follow-up tasks. The next phase for this run is in §3 (Recommended Next Steps)."
   },
   "columns": {
     "recordMeta": "Record",
@@ -39,7 +39,7 @@
     "stepwiseExecutionOrder": "Stepwise Execution Order"
   },
   "sectionIntro": {
-    "verdictCard": "At-a-glance verdict card — a summary that mirrors the values in `## 2. Final Verdict` and `## 6. Recommended Next Steps`.",
+    "verdictCard": "At-a-glance verdict card — a summary that mirrors the values in `## 7. Final Verdict` and `## 3. Recommended Next Steps`.",
     "clarificationCarryIn": "Unresolved `Clarification Items` from the prior report, re-examined against new evidence and carried in with updated status.",
     "ticketCoverage": "Summary of the core problems, requirements, and verification targets this run covered.",
     "executionStatus": "At-a-glance table of each worker's status, assigned model, and key findings.",
@@ -71,7 +71,7 @@
     "intro": "Final conclusion and recommended direction. `Direction` values: `continue-investigation / begin-implementation / approve / reject / hold`. When `task-type` is `final-verification`, `Verdict Token` is one of `accepted / conditional-accept / blocked` and serves as the `release-handoff` entry gate. For all other task-types: `not-applicable`."
   },
   "evidence": {
-    "sourceItemsColumnNote": "The `Source items` column is described in §1.1."
+    "sourceItemsColumnNote": "The `Source items` column is described in §6.1."
   },
   "roundHistory": {
     "round2SkippedReasonNote": "(one of: `queue-empty`, `max-rounds-1`, `all-reverify-non-result`, `not-skipped`, `convergence-disabled`, `single-analyser-only`)",

package/runtime/templates/reports/i18n/ko.json

@@ -15,8 +15,8 @@
     "declinedFixRecommendations": "- 없음.",
     "discrepancy": "- 없음.",
     "lingeringRisks": "- 추적 대상 잔존 위험 없음.",
-    "noClarification": "- 추가 정보 요청 없음. Section 2 의 최종 판단이 그대로 유효합니다.",
-    "noFollowUp": "- 후속 작업 없음. 본 run 의 다음 phase 는 §6 (Recommended Next Steps) 참고."
+    "noClarification": "- 추가 정보 요청 없음. Section 7 의 최종 판단이 그대로 유효합니다.",
+    "noFollowUp": "- 후속 작업 없음. 본 run 의 다음 phase 는 §3 (Recommended Next Steps) 참고."
   },
   "columns": {
     "recordMeta": "항목",
@@ -39,7 +39,7 @@
     "stepwiseExecutionOrder": "단계별 실행 순서"
   },
   "sectionIntro": {
-    "verdictCard": "한눈에 보는 결과 카드 — `## 2. Final Verdict` 와 `## 6. Recommended Next Steps` 의 값을 그대로 옮긴 요약입니다.",
+    "verdictCard": "한눈에 보는 결과 카드 — `## 7. Final Verdict` 와 `## 3. Recommended Next Steps` 의 값을 그대로 옮긴 요약입니다.",
     "clarificationCarryIn": "이전 보고서의 미해결 `Clarification Items` 를 새 증거로 재검토해 해소·갱신한 뒤 이어받은 항목입니다.",
     "ticketCoverage": "이 run 이 다룬 핵심 문제·요구사항·검증 대상 요약입니다.",
     "executionStatus": "각 worker 의 상태·배정 모델·핵심 finding 을 한눈에 보는 표입니다.",
@@ -71,7 +71,7 @@
     "intro": "최종 결론과 권장 방향입니다. `Direction` 값: `continue-investigation / begin-implementation / approve / reject / hold`. `task-type` 이 `final-verification` 일 때 `Verdict Token` 은 `accepted / conditional-accept / blocked` 중 하나이며 `release-handoff` 진입 게이트로 쓰입니다. 그 외 task-type 에서는 `not-applicable`."
   },
   "evidence": {
-    "sourceItemsColumnNote": "`Source items` 열 설명은 §1.1 과 동일합니다."
+    "sourceItemsColumnNote": "`Source items` 열 설명은 §6.1 과 동일합니다."
   },
   "roundHistory": {
     "round2SkippedReasonNote": "(허용 값: `queue-empty`, `max-rounds-1`, `all-reverify-non-result`, `not-skipped`, `convergence-disabled`, `single-analyser-only`)",

package/runtime/templates/reports/implementation-planning-input.template.md

@@ -108,7 +108,7 @@ The final report of an `implementation-planning` run MUST contain every section
 
 ## Stage Output Shape (reference)
 
-This run's final report MUST emit `## 4.5 Stage Map` and `## 4.5.<i> Stage <i>` sections per the implementation-planning profile §"Required deliverable shape". Two illustrative shapes:
+This run's final report MUST emit `## 5.5 Stage Map` and `## 5.5.<i> Stage <i>` sections per the implementation-planning profile §"Required deliverable shape". Two illustrative shapes:
 
 ### Shape A — single stage (small work)
 | stage | title | depends-on | step-count | exit-contract-summary |

package/runtime/templates/reports/release-handoff-input.template.md

@@ -26,7 +26,7 @@ taskType: "{{FM_TASK_TYPE}}"
 ## Source Verification Report
 
 - Path (project-relative) to the `final-verification` final-report whose verdict authorises this handoff:
-- Verbatim quoted `Verdict Token` row from that report's `## 2. Final Verdict` table (MUST have value `accepted`):
+- Verbatim quoted `Verdict Token` row from that report's `## 7. Final Verdict` table (MUST have value `accepted`):
 - Run timestamp of that final-verification run:
 
 > If this section is empty or cites a `Verdict Token` value other than `accepted`, the lead MUST end the run immediately and route back to `final-verification`. Release-handoff never operates on `conditional-accept` or `blocked` outcomes.

package/runtime/templates/reports/user-response.template.md

@@ -18,7 +18,7 @@ created-at: <ISO 8601 UTC timestamp>
 
 ## Body 스키마
 
-본문은 `# User Response` 단일 H1 헤딩 아래, 각 응답을 `## <Response ID>` 헤딩으로 구분합니다. Response ID 는 final-report `## 5. Clarification Items` 표의 `ID` 컬럼(`C-NNN`)을 그대로 인용합니다.
+본문은 `# User Response` 단일 H1 헤딩 아래, 각 응답을 `## <Response ID>` 헤딩으로 구분합니다. Response ID 는 final-report `## 1. Clarification Items` 표의 `ID` 컬럼(`C-NNN`)을 그대로 인용합니다.
 
 각 응답 블록의 schema:
 

package/runtime/templates/worker-prompt-preamble.md

@@ -8,7 +8,7 @@ It replaces the previous practice of inlining ~80 lines of identical boilerplate
 
 ## Required reading (analysis workers + report-writer worker)
 
-You are required to read every input file enumerated by the dispatcher (the lead's prompt lists them under `[Required reading]`) from the very first character to the very last character before you produce any analysis output. Skimming, partial reads, jumping to a single section, or relying on prior knowledge of a similar file's structure is not acceptable. Each file may contain decisive context that is not surfaced in its summary or first page.
+You are required to read every primary input file enumerated by the dispatcher (the lead's prompt lists them under `[Required reading]`) from the very first character to the very last character before you produce any analysis output. Skimming, partial reads, jumping to a single section, or relying on prior knowledge of a similar file's structure is not acceptable. Source files listed as fallback/evidence paths are read on demand when you need to verify a citation, resolve ambiguity, or inspect material the packet says it omitted.
 
 ### Audience-scoped enumeration (BLOCKING — performance optimization)
 
@@ -16,16 +16,16 @@ Different recipients need different files. Do NOT include `final-report-template
 
 | Recipient | Files included in `[Required reading]` |
 |---|---|
-| Claude / Codex / Gemini analysis workers | task-brief, analysis-profile, analysis-material (if present), reference-expectations, clarification-response (if carry-in) |
+| Claude / Codex / Gemini analysis workers | analysis-packet.md as the primary compact input; task-brief, analysis-profile, analysis-material, reference-expectations, and clarification-response remain source/fallback paths, not automatic first-read files |
 | Report writer worker (Phase 6) | all of the above **plus** the instruction-set-local `final-report-template.md` (phase-stripped) and `final-report-schema.json` (per-task-type excerpt) — NOT the full `templates/reports/...` / `schemas/...` sources |
 | Reverify dispatches (Phase 5.5, lightweight mode) | **do NOT inject `[Required reading]` at all** — see [okstra-convergence](../skills/okstra-convergence/SKILL.md) "Reverify prompt: required-reading suppression". |
 
 ### Reading rules
 
 - Use a single `Read` tool call per file with no `offset` and no `limit`. If a file is genuinely too large for one read, page through with explicit `offset` / `limit` covering the entire file, and state the page boundaries in your Findings.
-- For the carry-in clarification response, walk every row of `## 5. Clarification Items` (`C-001`, `C-002`, ...) in full, including rows whose `User input` cell is blank — a blank `User input` with `Status=open` is itself a signal you must surface. The structural similarity between the prior final report and the upcoming output is NOT a license to skim.
+- For the carry-in clarification response, walk every row of `## 1. Clarification Items` (`C-001`, `C-002`, ...) in full, including rows whose `User input` cell is blank — a blank `User input` with `Status=open` is itself a signal you must surface. The structural similarity between the prior final report and the upcoming output is NOT a license to skim.
 - Write the Reading Confirmation block to your **audit sidecar** at `runs/<task-type>/worker-results/<worker>-audit-<task-type>-<seq>.md` (sibling to the main worker-results file). One short line per input file confirming end-to-end reading. Do NOT include a `## 0. Reading Confirmation` heading in the main worker-results file — the validator fails worker-results that contain one. If you cannot truthfully confirm a file end-to-end, record a `tool-failure` in the errors sidecar instead of fabricating Findings.
-- Do not collapse multiple input files into a single mental summary before reading them all individually. Each file has its own canonical role: brief = the user's request, profile = the lead's rules for this phase, reference-expectations = ground-truth config/deployment values, clarification-response = prior run's open questions and the user's answers, final-report-template = the structure your eventual writeup must conform to. Conflating them loses signal.
+- Treat `analysis-packet.md` as the canonical primary analysis input. It preserves the source files' frontmatter and extracts the task-specific brief, phase focus, reference expectations, and carry-in clarification rows. If the packet appears incomplete or a finding depends on a source citation, open the corresponding source file and cite it directly.
 
 ---
 

package/runtime/validators/lib/fixtures.sh

@@ -350,13 +350,13 @@ report_lines.extend(
         "| **전체 합계** | **`2`** | **`2`** | **`$0.02`** |",
         "| Codex/Gemini CLI 추가 비용 |  |  | `$0.00` |",
         "",
-        "## 2. Final Verdict",
+        "## 7. Final Verdict",
         "",
         "| 항목 | 값 |",
         "|------|----|",
         "| Verdict Token | `accepted` |",
         "",
-        "## 4.8 Final Verification Deliverables",
+        "## 5.8 Final Verification Deliverables",
         "",
         "Source Implementation Report / Acceptance Blockers / Residual Risk / "
         "Validation Evidence / Read-only Command Log / Conditional Acceptance "

package/runtime/validators/validate-implementation-plan-stages.py

@@ -12,9 +12,9 @@ from dataclasses import dataclass
 from pathlib import Path
 from typing import List, Tuple
 
-STAGE_MAP_HEADING = re.compile(r"^##\s+4\.5\s+Stage\s+Map\b", re.M)
+STAGE_MAP_HEADING = re.compile(r"^##\s+5\.5\s+Stage\s+Map\b", re.M)
 STAGE_SECTION = re.compile(
-    r"^##\s+4\.5\.(\d+)\s+Stage\s+\1\s*:\s*(.+)$", re.M
+    r"^##\s+5\.5\.(\d+)\s+Stage\s+\1\s*:\s*(.+)$", re.M
 )
 REQUIRED_SUBSECTIONS = (
     "Carry-In",
@@ -48,7 +48,7 @@ class ValidationError:
 def _check_stage_map_present(text: str) -> List[ValidationError]:
     if not STAGE_MAP_HEADING.search(text):
         return [ValidationError("S1", 0,
-            "section '## 4.5 Stage Map' is missing")]
+            "section '## 5.5 Stage Map' is missing")]
     return []
 
 
@@ -91,15 +91,15 @@ def _parse_stage_map(text: str) -> Tuple[List[StageMeta], List[ValidationError]]
 
 
 def _slice_stage_section(text: str, stage_number: int) -> str:
-    """Return the body of `## 4.5.<n> Stage <n>:` up to the next stage heading."""
+    """Return the body of `## 5.5.<n> Stage <n>:` up to the next stage heading."""
     start_m = re.search(
-        rf"^##\s+4\.5\.{stage_number}\s+Stage\s+{stage_number}\s*:", text, re.M
+        rf"^##\s+5\.5\.{stage_number}\s+Stage\s+{stage_number}\s*:", text, re.M
     )
     if not start_m:
         return ""
     start = start_m.end()
     nxt = re.search(
-        rf"^##\s+4\.5\.{stage_number + 1}\s+Stage\s+", text[start:], re.M
+        rf"^##\s+5\.5\.{stage_number + 1}\s+Stage\s+", text[start:], re.M
     )
     return text[start: start + nxt.start()] if nxt else text[start:]
 
@@ -134,10 +134,10 @@ def _check_each_stage_section(text: str, stages: List[StageMeta]) -> List[Valida
     errs: List[ValidationError] = []
     for s in stages:
         if not re.search(
-            rf"^##\s+4\.5\.{s.stage_number}\s+Stage\s+{s.stage_number}\s*:", text, re.M
+            rf"^##\s+5\.5\.{s.stage_number}\s+Stage\s+{s.stage_number}\s*:", text, re.M
         ):
             errs.append(ValidationError("S3", s.stage_number,
-                f"stage section '## 4.5.{s.stage_number} Stage {s.stage_number}:' missing"))
+                f"stage section '## 5.5.{s.stage_number} Stage {s.stage_number}:' missing"))
             continue
         section = _slice_stage_section(text, s.stage_number)
 
@@ -231,7 +231,7 @@ def _check_parallel_safety(text: str, stages: List[StageMeta]) -> List[Validatio
 def collect_validation_errors(text: str) -> List[ValidationError]:
     """All S1–S9 checks against the report text; empty list means valid.
 
-    S1 (missing `## 4.5 Stage Map` heading) makes the rest unparseable, so it
+    S1 (missing `## 5.5 Stage Map` heading) makes the rest unparseable, so it
     short-circuits. Shared by `main()` (CLI / implementation entry) and the
     implementation-planning boundary check in validate-run.py — one validator,
     one reference point, enforced at both produce-time and consume-time."""

package/runtime/validators/validate-report-views.py

@@ -6,12 +6,12 @@ Checks, for a given final-report MD path:
   1. ``*.html`` sibling exists.
   2. HTML's ``source-sha256`` in run-meta matches the current MD body —
      stale html detection.
-  3. HTML's §4.6 / §4.7 / §4.8 deliverable regions contain no
+  3. HTML's §5.6 / §5.7 / §5.8 deliverable regions contain no
      ``<textarea>`` / ``<input>`` / ``<select>`` (form-attach is
-     restricted to §5 clarification rows).
+     restricted to §1 clarification rows).
   4. HTML has no external URLs in ``<script src=>`` / ``<link href=>`` /
      ``<img src=>`` — self-contained guarantee.
-  5. Every Response ID in HTML matches the §5 Clarification Items table
+  5. Every Response ID in HTML matches the §1 Clarification Items table
      of the source MD (1:1).
 
 Exit codes: 0 on success, 1 on any failure. Failures are printed one
@@ -54,7 +54,7 @@ def _main_body(html_text: str) -> str:
 
 def _no_form_sections(html_body: str) -> list[str]:
     """Return a list of strings, each being the rendered chunk of a
-    no-form section (4.6 / 4.7 / 4.8) up to the next h2/h3. Used to
+    no-form section (5.6 / 5.7 / 5.8) up to the next h2/h3. Used to
     check that no form controls live inside them.
     """
     chunks: list[str] = []
@@ -65,7 +65,7 @@ def _no_form_sections(html_body: str) -> list[str]:
     ]
     for i, m in enumerate(headings):
         text = m.group(2).strip()
-        if not text.startswith(("4.6", "4.7", "4.8")):
+        if not text.startswith(("5.6", "5.7", "5.8")):
             continue
         start = m.end()
         end = headings[i + 1].start() if i + 1 < len(headings) else len(html_body)
@@ -82,9 +82,9 @@ def validate(report_path: Path) -> list[str]:
     html_path = report_path.with_name(report_path.stem + ".html")
     md_ids = _md_response_ids(md)
 
-    # (1) sibling artifact exists — conditional on §5 clarification rows.
+    # (1) sibling artifact exists — conditional on §1 clarification rows.
     # The html view's sole value over the MD is its embedded form widgets
-    # for §5 C-* rows, so a clarification-free report intentionally has no
+    # for §1 C-* rows, so a clarification-free report intentionally has no
     # html sibling (see report_views.report_has_clarification_items). When
     # there are no C-* rows and no html, that is the expected skip state.
     if not html_path.is_file():
@@ -115,7 +115,7 @@ def validate(report_path: Path) -> list[str]:
     for chunk in _no_form_sections(html_body):
         if "<textarea" in chunk or "<input" in chunk or "<select" in chunk:
             failures.append(
-                "html §4.6/§4.7/§4.8 deliverable section contains a form control"
+                "html §5.6/§5.7/§5.8 deliverable section contains a form control"
             )
             break
 
@@ -123,14 +123,14 @@ def validate(report_path: Path) -> list[str]:
     if _EXTERNAL_URL_RE.search(html_text):
         failures.append("html contains external URL in script/link/img — must be self-contained")
 
-    # (5) Response ID parity: HTML form rows ↔ §5 C-* rows in MD.
+    # (5) Response ID parity: HTML form rows ↔ §1 C-* rows in MD.
     # Bidirectional — catches both "MD has C-* the HTML lost" AND
     # "HTML has stale C-* that the current MD no longer declares".
     # (md_ids computed once at the top.)
     html_ids = sorted(set(_RESPONSE_ID_ATTR_RE.findall(html_text)))
     if md_ids != html_ids:
         failures.append(
-            f"Response ID mismatch: MD §5 has {md_ids}, HTML has {html_ids}"
+            f"Response ID mismatch: MD §1 has {md_ids}, HTML has {html_ids}"
         )
 
     return failures

package/runtime/validators/validate-run.py

@@ -644,32 +644,32 @@ _DEPRECATED_FINAL_REPORT_PATTERNS: tuple[tuple[re.Pattern, str], ...] = (
         "the YAML frontmatter `approved: true|false` field. Delete the body section.",
     ),
     (
-        re.compile(r"^###[ \t]+4\.5\.8[ \t]+User Approval Request\b", re.MULTILINE),
-        "deprecated `### 4.5.8 User Approval Request` stub — approval gate moved "
-        "to the YAML frontmatter `approved: true|false` field. Delete the §4.5.8 heading + body.",
+        re.compile(r"^###[ \t]+5\.5\.8[ \t]+User Approval Request\b", re.MULTILINE),
+        "deprecated `### 5.5.8 User Approval Request` stub — approval gate moved "
+        "to the YAML frontmatter `approved: true|false` field. Delete the §5.5.8 heading + body.",
     ),
     (
-        re.compile(r"^###[ \t]+4\.5\.9[ \t]+Open Questions\b", re.MULTILINE),
-        "deprecated `### 4.5.9 Open Questions` block — promote each row into "
-        "`## 5. Clarification Items` with `Kind=decision` (and `Blocks=approval` "
+        re.compile(r"^###[ \t]+5\.5\.9[ \t]+Open Questions\b", re.MULTILINE),
+        "deprecated `### 5.5.9 Open Questions` block — promote each row into "
+        "`## 1. Clarification Items` with `Kind=decision` (and `Blocks=approval` "
         "if it gates the frontmatter approval flag).",
     ),
     (
         re.compile(
-            r"^###[ \t]+5\.1[ \t]+(?:추가 자료 요청|Additional Materials)\b",
+            r"^###[ \t]+1\.1[ \t]+(?:추가 자료 요청|Additional Materials)\b",
             re.MULTILINE,
         ),
-        "deprecated `### 5.1 추가 자료 요청` / `Additional Materials` sub-section — "
-        "every clarification item lives as one row of the unified `## 5. "
+        "deprecated `### 1.1 추가 자료 요청` / `Additional Materials` sub-section — "
+        "every clarification item lives as one row of the unified `## 1. "
         "Clarification Items` 8-column table (`Kind=material`).",
     ),
     (
         re.compile(
-            r"^###[ \t]+5\.2[ \t]+(?:사용자 확인 질문|Questions for the User)\b",
+            r"^###[ \t]+1\.2[ \t]+(?:사용자 확인 질문|Questions for the User)\b",
             re.MULTILINE,
         ),
-        "deprecated `### 5.2 사용자 확인 질문` / `Questions for the User` "
-        "sub-section — collapse into the unified `## 5. Clarification Items` "
+        "deprecated `### 1.2 사용자 확인 질문` / `Questions for the User` "
+        "sub-section — collapse into the unified `## 1. Clarification Items` "
         "8-column table (`Kind=decision`).",
     ),
 )
@@ -708,7 +708,7 @@ def validate_report(
             "final report is missing the top-of-report `## Verdict Card` block — "
             "render it between the report header and the (conditional) Approval "
             "block. Its Verdict Token / Direction / Next Step cells must byte-match "
-            "the corresponding cells in `## 2. Final Verdict` and `## 6.` first item."
+            "the corresponding cells in `## 7. Final Verdict` and `## 3.` first item."
         )
 
     # Reading Confirmation belongs in the worker audit sidecar, not the
@@ -860,7 +860,7 @@ PLANNING_REQUIRED_SECTIONS = (
     "Plan Body Verification",
 )
 
-# §4.7 implementation deliverables — substring scan against report body.
+# §5.7 implementation deliverables — substring scan against report body.
 IMPLEMENTATION_REQUIRED_SECTIONS = (
     "Approved Plan Reference",
     "Commit List",
@@ -872,7 +872,7 @@ IMPLEMENTATION_REQUIRED_SECTIONS = (
     "Routing Recommendation",
 )
 
-# §4.8 final-verification deliverables — substring scan against report body.
+# §5.8 final-verification deliverables — substring scan against report body.
 FINAL_VERIFICATION_REQUIRED_SECTIONS = (
     "Source Implementation Report",
     "Acceptance Blockers",
@@ -892,7 +892,7 @@ FINAL_VERIFICATION_VERDICT_TOKENS = (
     "blocked",
 )
 
-# `## 2. Final Verdict` Verdict Token cell — captures the value between
+# `## 7. Final Verdict` Verdict Token cell — captures the value between
 # backticks on the `Verdict Token` row. Tolerant to extra column whitespace
 # and to leading bold/italic markers in the label cell.
 _FINAL_VERDICT_TOKEN_RE = re.compile(
@@ -909,20 +909,20 @@ _VERDICT_CARD_BLOCK_RE = re.compile(
     re.DOTALL | re.MULTILINE,
 )
 
-# `## 2. Final Verdict` block scope — used to scope the Verdict Token
+# `## 7. Final Verdict` block scope — used to scope the Verdict Token
 # regex so that we don't accidentally match a Verdict Token row that
 # lives in the Verdict Card or anywhere else.
 _FINAL_VERDICT_BLOCK_RE = re.compile(
-    r"^##[ \t]+2\.[ \t]+Final Verdict[ \t]*$\n(?P<body>.*?)(?=^##[ \t]|\Z)",
+    r"^##[ \t]+7\.[ \t]+Final Verdict[ \t]*$\n(?P<body>.*?)(?=^##[ \t]|\Z)",
     re.DOTALL | re.MULTILINE,
 )
 
-# `## 4.6 Release Handoff Deliverables` and `## 4.6.6 Merge Conflict
+# `## 5.6 Release Handoff Deliverables` and `## 5.6.6 Merge Conflict
 # Probe` are required when task-type == release-handoff. The probe sub-
 # section was retro-added to the template; old runs that predate it ship
 # without it, but new runs must include it.
 _MERGE_CONFLICT_PROBE_HEADING_RE = re.compile(
-    r"^###[ \t]+4\.6\.6[ \t]+Merge Conflict Probe\b", re.MULTILINE
+    r"^###[ \t]+5\.6\.6[ \t]+Merge Conflict Probe\b", re.MULTILINE
 )
 
 PLAN_VERIFY_GATE_VALUES = (
@@ -932,7 +932,7 @@ PLAN_VERIFY_GATE_VALUES = (
     "aborted-non-result",
 )
 
-# `Gate result:` line in §4.5.9 of the final report — captures the value
+# `Gate result:` line in §5.5.9 of the final report — captures the value
 # token that follows. Tolerates arbitrary markdown formatting between the
 # label and the value (backticks for inline code, double-asterisks for
 # bold, colons, hyphens, whitespace). The captured value is then
@@ -954,8 +954,8 @@ _FRONTMATTER_BLOCK_RE = re.compile(r"\A---\n(.*?)\n---\n", re.DOTALL)
 
 
 def _extract_final_verdict_token(content: str) -> str | None:
-    """Return the `Verdict Token` cell value from the `## 2. Final Verdict`
-    block, or None when the row is absent. Scoped to §2 so the Verdict
+    """Return the `Verdict Token` cell value from the `## 7. Final Verdict`
+    block, or None when the row is absent. Scoped to §7 so the Verdict
     Card row (which has the same shape) does not shadow the authoritative
     value.
     """
@@ -980,7 +980,7 @@ def _extract_verdict_card_token(content: str) -> str | None:
 
 
 def _validate_verdict_card_consistency(content: str, failures: list[str]) -> None:
-    """Verdict Card is a non-authoritative index of §2. If both blocks
+    """Verdict Card is a non-authoritative index of §7. If both blocks
     carry a Verdict Token row, the values MUST byte-match (modulo case
     and surrounding whitespace) — divergence is a contract violation per
     `okstra-report-writer` SKILL.md "Authoring Contract".
@@ -988,15 +988,15 @@ def _validate_verdict_card_consistency(content: str, failures: list[str]) -> Non
     card_value = _extract_verdict_card_token(content)
     final_value = _extract_final_verdict_token(content)
     if card_value is None or final_value is None:
-        # Missing-Card and missing-§2 are surfaced by other checks; this
+        # Missing-Card and missing-§7 are surfaced by other checks; this
         # function only enforces the consistency contract between the two.
         return
     if card_value.strip().lower() != final_value.strip().lower():
         failures.append(
             "Verdict Card `Verdict Token` value "
-            f"`{card_value}` does not match `## 2. Final Verdict` value "
+            f"`{card_value}` does not match `## 7. Final Verdict` value "
             f"`{final_value}` — the Card is a non-authoritative index and "
-            "MUST byte-match §2. Either fix the Card or update §2; do not "
+            "MUST byte-match §7. Either fix the Card or update §7; do not "
             "ship divergent values."
         )
 
@@ -1161,7 +1161,7 @@ def _load_stage_validator():
 
 def _append_stage_structure_failures(content: str, failures: list[str]) -> None:
     """Enforce the Stage Map structural contract at the implementation-planning
-    boundary. Without this, a plan missing `## 4.5 Stage Map` passes the
+    boundary. Without this, a plan missing `## 5.5 Stage Map` passes the
     planning gate, gets approved, and only fails later at the `implementation`
     entry (validators/validate-implementation-plan-stages.py via
     prepare_task_bundle). Running the same validator here moves the failure to
@@ -1189,7 +1189,7 @@ def validate_phase_boundary(
     skipped its core outputs (or an implementation run that ran under the
     wrong task type).
 
-    Additionally enforces the Plan Body Verification gate (§4.5.9):
+    Additionally enforces the Plan Body Verification gate (§5.5.9):
     - gate ∈ {passed, passed-with-dissent} → top-level Approval checkbox
       MUST be present.
     - gate ∈ {blocked-by-disagreement, aborted-non-result} → checkbox
@@ -1200,7 +1200,7 @@ def validate_phase_boundary(
         return
     content = report_path.read_text()
 
-    # Verdict Card vs §2. Final Verdict Verdict Token consistency. The Card
+    # Verdict Card vs §7. Final Verdict Verdict Token consistency. The Card
     # is a non-authoritative index; divergence is a contract violation.
     _validate_verdict_card_consistency(content, failures)
 
@@ -1208,7 +1208,7 @@ def validate_phase_boundary(
         for needle in IMPLEMENTATION_REQUIRED_SECTIONS:
             if needle not in content:
                 failures.append(
-                    "implementation report is missing required §4.7 "
+                    "implementation report is missing required §5.7 "
                     f"deliverable section: `{needle}`"
                 )
 
@@ -1216,13 +1216,13 @@ def validate_phase_boundary(
         for needle in FINAL_VERIFICATION_REQUIRED_SECTIONS:
             if needle not in content:
                 failures.append(
-                    "final-verification report is missing required §4.8 "
+                    "final-verification report is missing required §5.8 "
                     f"deliverable section: `{needle}`"
                 )
         token_value = _extract_final_verdict_token(content)
         if token_value is None:
             failures.append(
-                "final-verification report `## 2. Final Verdict` table is "
+                "final-verification report `## 7. Final Verdict` table is "
                 "missing the `Verdict Token` row — required by the release-"
                 "handoff entry gate."
             )
@@ -1236,7 +1236,7 @@ def validate_phase_boundary(
     if task_type == "release-handoff":
         if _MERGE_CONFLICT_PROBE_HEADING_RE.search(content) is None:
             failures.append(
-                "release-handoff report is missing `### 4.6.6 Merge Conflict "
+                "release-handoff report is missing `### 5.6.6 Merge Conflict "
                 "Probe` sub-section — required by the release-handoff profile "
                 "(self-review 6, merge-conflict probe audit). When the run is "
                 "`local only` / `skip`, record the single line `- Not run "
@@ -1260,7 +1260,7 @@ def validate_phase_boundary(
         if "Plan Body Verification" in content:
             failures.append(
                 "implementation-planning report has `Plan Body Verification` "
-                "section but no `Gate result:` line — required by §4.5.9."
+                "section but no `Gate result:` line — required by §5.5.9."
             )
         return
     gate_value = gate_match.group("value").strip().lower()
@@ -1655,7 +1655,7 @@ def _rerender_report_views_after_autofix(report_path: Path) -> str:
     except Exception as exc:  # noqa: BLE001
         return f"report-views re-render failed: {exc}"
     if rendered is None:
-        return "report-views skipped (no §5 clarification rows)"
+        return "report-views skipped (no §1 clarification rows)"
     return "report-views re-rendered"
 
 

package/runtime/validators/validate_improvement_report.py

@@ -205,9 +205,9 @@ def _check_candidates_table(
     body: str, brief_frontmatter: dict, errors: list[str]
 ) -> None:
     """Items 1–7 coordinator — parse table, check header, delegate cap and rows."""
-    rows = _read_section_table(body, "4.9 Improvement Candidates")
+    rows = _read_section_table(body, "5.9 Improvement Candidates")
     if not rows:
-        errors.append("missing or empty `## 4.9 Improvement Candidates` table")
+        errors.append("missing or empty `## 5.9 Improvement Candidates` table")
         return
     header, *data = rows
     expected_columns = [
@@ -216,7 +216,7 @@ def _check_candidates_table(
     ]
     if header != expected_columns:
         errors.append(
-            f"`## 4.9 Improvement Candidates` header must be {expected_columns}, "
+            f"`## 5.9 Improvement Candidates` header must be {expected_columns}, "
             f"got {header}"
         )
     _check_candidates_cap(len(data), brief_frontmatter, errors)
@@ -228,18 +228,18 @@ def _check_candidates_table(
 def _check_final_verdict(
     body: str, errors: list[str]
 ) -> list[list[str]]:
-    """Item 8 — ## 2. Final Verdict verdict token is in enum.
+    """Item 8 — ## 7. Final Verdict verdict token is in enum.
 
     Returns the parsed final_verdict_rows (empty list when absent).
     """
-    final_verdict_rows = _read_section_table(body, "2. Final Verdict")
+    final_verdict_rows = _read_section_table(body, "7. Final Verdict")
     if not final_verdict_rows:
-        errors.append("missing `## 2. Final Verdict` block")
+        errors.append("missing `## 7. Final Verdict` block")
         return []
     token_cell = final_verdict_rows[-1][0] if final_verdict_rows[-1] else ""
     if token_cell not in _VERDICT_TOKENS:
         errors.append(
-            f"`## 2. Final Verdict` Verdict Token '{token_cell}' must be one of {_VERDICT_TOKENS}"
+            f"`## 7. Final Verdict` Verdict Token '{token_cell}' must be one of {_VERDICT_TOKENS}"
         )
     return final_verdict_rows
 
@@ -253,7 +253,7 @@ def _check_verdict_card(
         errors.append("missing `## Verdict Card` block")
         return
     if final_verdict_rows and verdict_card_rows[-1] != final_verdict_rows[-1]:
-        errors.append("Verdict Card row must byte-match `## 2. Final Verdict` row")
+        errors.append("Verdict Card row must byte-match `## 7. Final Verdict` row")
 
 
 def validate_improvement_report(

package/src/_python-helper.mjs

@@ -1,12 +1,12 @@
 import { spawn } from "node:child_process";
-import { resolvePaths } from "./paths.mjs";
+import { buildPythonpath, resolvePaths } from "./paths.mjs";
 
 export async function runPythonSnippet({ script, args = [], extraEnv = {} }) {
   const paths = await resolvePaths();
   return new Promise((resolve) => {
     const child = spawn("python3", ["-c", script, ...args], {
       stdio: ["ignore", "pipe", "pipe"],
-      env: { ...process.env, PYTHONPATH: paths.pythonpath, ...extraEnv },
+      env: { ...process.env, PYTHONPATH: buildPythonpath(paths), ...extraEnv },
     });
     let stdout = "";
     let stderr = "";
@@ -22,7 +22,7 @@ export async function runPythonModule({ module, args = [], extraEnv = {}, stdio
   return new Promise((resolve) => {
     const child = spawn("python3", ["-m", module, ...args], {
       stdio: stdio === "capture" ? ["ignore", "pipe", "pipe"] : ["ignore", "inherit", "inherit"],
-      env: { ...process.env, PYTHONPATH: paths.pythonpath, ...extraEnv },
+      env: { ...process.env, PYTHONPATH: buildPythonpath(paths), ...extraEnv },
     });
     let stdout = "";
     let stderr = "";

package/src/context-cost.mjs

@@ -0,0 +1,27 @@
+import { runPythonModule } from "./_python-helper.mjs";
+
+const USAGE = `okstra context-cost — estimate context/read cost for a task bundle
+
+Usage:
+  okstra context-cost <task-root>
+  okstra context-cost <task-key> --project-root <dir>
+  okstra context-cost <task-key> --cwd <dir>
+
+Output: JSON with task file counts, current-run counts, lead Phase 1 read
+surface, analysis-worker initial read surface, and report-writer synthesis
+surface. This is read-only; it never mutates task artifacts.
+`;
+
+export async function run(args) {
+  if (args.includes("--help") || args.includes("-h")) {
+    process.stdout.write(USAGE);
+    return 0;
+  }
+
+  const result = await runPythonModule({
+    module: "okstra_ctl.context_cost",
+    args,
+    stdio: "inherit-stdout",
+  });
+  return result.code ?? 0;
+}

package/src/install.mjs

@@ -22,6 +22,7 @@ const BIN_ENTRYPOINTS = [
   "okstra-codex-exec.sh",
   "okstra-gemini-exec.sh",
   "okstra-trace-cleanup.sh",
+  "okstra-team-reconcile.sh",
   "okstra-central.sh",
   "okstra-token-usage.py",
   "okstra-error-log.py",

package/src/uninstall.mjs

@@ -8,6 +8,7 @@ const BIN_ENTRYPOINTS = [
   "okstra-codex-exec.sh",
   "okstra-gemini-exec.sh",
   "okstra-trace-cleanup.sh",
+  "okstra-team-reconcile.sh",
   "okstra-central.sh",
   "okstra-token-usage.py",
   "okstra-error-log.py",