maestro-flow 0.4.20 → 0.4.22

package/.agents/skills/team-swarm/scripts/aco.py

@@ -89,7 +89,7 @@ def cmd_init(args: argparse.Namespace) -> None:
     paths = SessionPaths(Path(args.session))
     if not paths.config.exists():
         _fail(2, f"config not found: {paths.config}")
-    config = json.loads(paths.config.read_text())
+    config = json.loads(paths.config.read_text(encoding="utf-8"))
 
     paths.ensure_dirs()
 
@@ -107,7 +107,7 @@ def cmd_init(args: argparse.Namespace) -> None:
         "start_nodes": config.get("task_space", {}).get("start_nodes", "any"),
         "edges": config.get("task_space", {}).get("edges", "complete"),
     }
-    paths.task_space.write_text(json.dumps(task_space, indent=2, ensure_ascii=False))
+    paths.task_space.write_text(json.dumps(task_space, indent=2, ensure_ascii=False), encoding="utf-8")
 
     # Initialize pheromone
     aco_cfg = config.get("aco", {})
@@ -148,9 +148,9 @@ def _pick_start_node(nodes: List[str], state: PheromoneState, mode: str) -> str:
 
 def cmd_select(args: argparse.Namespace) -> None:
     paths = SessionPaths(Path(args.session))
-    config = json.loads(paths.config.read_text())
+    config = json.loads(paths.config.read_text(encoding="utf-8"))
     state = PheromoneState.load(paths.pheromone_current)
-    task_space = json.loads(paths.task_space.read_text())
+    task_space = json.loads(paths.task_space.read_text(encoding="utf-8"))
 
     n_ants = config.get("swarm", {}).get("n_ants", 5)
     nodes = task_space["nodes"]
@@ -190,7 +190,7 @@ def _load_iteration_artifacts(paths: SessionPaths, iteration: int) -> List[dict]
     artifacts = []
     for f in files:
         try:
-            artifacts.append(json.loads(Path(f).read_text()))
+            artifacts.append(json.loads(Path(f).read_text(encoding="utf-8")))
         except json.JSONDecodeError as e:
             print(f"warning: skipped malformed artifact {f}: {e}", file=sys.stderr)
     return artifacts
@@ -213,9 +213,9 @@ def _validate_artifact(art: dict, valid_nodes: set) -> Optional[str]:
 
 def cmd_update(args: argparse.Namespace) -> None:
     paths = SessionPaths(Path(args.session))
-    config = json.loads(paths.config.read_text())
+    config = json.loads(paths.config.read_text(encoding="utf-8"))
     state = PheromoneState.load(paths.pheromone_current)
-    task_space = json.loads(paths.task_space.read_text())
+    task_space = json.loads(paths.task_space.read_text(encoding="utf-8"))
     valid_nodes = set(task_space["nodes"])
 
     artifacts = _load_iteration_artifacts(paths, args.iter)
@@ -266,7 +266,7 @@ def cmd_update(args: argparse.Namespace) -> None:
     # Elitist: re-load best history, deposit extra on best path
     best_data = None
     if paths.best.exists():
-        best_data = json.loads(paths.best.read_text())
+        best_data = json.loads(paths.best.read_text(encoding="utf-8"))
     current_best = max(scored, key=lambda x: x["score"]) if scored else None
     if current_best:
         best_art = next(a for a in artifacts if a["ant_id"] == current_best["ant_id"])
@@ -281,7 +281,7 @@ def cmd_update(args: argparse.Namespace) -> None:
                 "evidence": best_art.get("evidence", []),
                 "updated_at": time.time(),
             }
-            paths.best.write_text(json.dumps(best_data, indent=2, ensure_ascii=False))
+            paths.best.write_text(json.dumps(best_data, indent=2, ensure_ascii=False), encoding="utf-8")
         # Elite deposit
         state.deposit(best_data["path"], best_data["score"])
 
@@ -292,14 +292,14 @@ def cmd_update(args: argparse.Namespace) -> None:
 
     # Persist trails
     trails_file = paths.trails / f"{args.iter}.jsonl"
-    trails_file.write_text("\n".join(json.dumps(t, ensure_ascii=False) for t in trail_log))
+    trails_file.write_text("\n".join(json.dumps(t, ensure_ascii=False) for t in trail_log), encoding="utf-8")
 
     mean_score = sum(s["score"] for s in scored) / len(scored) if scored else 0.0
     best_score = best_data["score"] if best_data else 0.0
     prev_best = 0.0
     history_files = sorted(paths.pheromone_history.glob("*.json"))
     if len(history_files) >= 2:
-        prev = json.loads(history_files[-2].read_text())
+        prev = json.loads(history_files[-2].read_text(encoding="utf-8"))
         prev_best = prev.get("stats", {}).get("best_known", best_score)
     delta = best_score - prev_best
 
@@ -323,7 +323,7 @@ def cmd_update(args: argparse.Namespace) -> None:
 
 def cmd_converged(args: argparse.Namespace) -> None:
     paths = SessionPaths(Path(args.session))
-    config = json.loads(paths.config.read_text())
+    config = json.loads(paths.config.read_text(encoding="utf-8"))
     cv = config.get("convergence", {})
 
     state = PheromoneState.load(paths.pheromone_current)
@@ -339,7 +339,7 @@ def cmd_converged(args: argparse.Namespace) -> None:
     }
 
     if paths.best.exists():
-        metrics["best_score"] = json.loads(paths.best.read_text()).get("score", 0.0)
+        metrics["best_score"] = json.loads(paths.best.read_text(encoding="utf-8")).get("score", 0.0)
 
     # max_iterations
     max_iter = cv.get("max_iterations", 5)
@@ -397,7 +397,7 @@ def cmd_report(args: argparse.Namespace) -> None:
 
     best = None
     if paths.best.exists():
-        best = json.loads(paths.best.read_text())
+        best = json.loads(paths.best.read_text(encoding="utf-8"))
 
     # Top-K trails across all iterations
     all_trails = []
@@ -410,7 +410,7 @@ def cmd_report(args: argparse.Namespace) -> None:
     # Convergence curve
     curve = []
     for hf in sorted(paths.pheromone_history.glob("*.json"), key=lambda p: int(p.stem)):
-        snap = json.loads(hf.read_text())
+        snap = json.loads(hf.read_text(encoding="utf-8"))
         curve.append({
             "iteration": snap["iteration"],
             "entropy": snap["stats"]["entropy"],

package/.agents/skills/team-swarm/scripts/pheromone.py

@@ -114,11 +114,11 @@ class PheromoneState:
 
     def save(self, path: Path) -> None:
         path.parent.mkdir(parents=True, exist_ok=True)
-        path.write_text(json.dumps(self.to_dict(), indent=2, ensure_ascii=False))
+        path.write_text(json.dumps(self.to_dict(), indent=2, ensure_ascii=False), encoding="utf-8")
 
     @classmethod
     def load(cls, path: Path) -> "PheromoneState":
-        return cls.from_dict(json.loads(path.read_text()))
+        return cls.from_dict(json.loads(path.read_text(encoding="utf-8")))
 
     def select_neighbors(
         self,

package/.agents/skills/team-swarm/scripts/scoring.py

@@ -62,7 +62,7 @@ def load_verified_scores(scores_file: Path) -> Dict[str, float]:
     """Load pre-computed verified_scores from scorer role output (if exists)."""
     if not scores_file.exists():
         return {}
-    data = json.loads(scores_file.read_text())
+    data = json.loads(scores_file.read_text(encoding="utf-8"))
     return {
         ant_id: entry["verified_score"]
         for ant_id, entry in data.get("scores", {}).items()

package/.agy/skills/maestro-ralph-execute/SKILL.md

@@ -50,6 +50,7 @@ HARD RULES:
 6. **Self-invocation chain** — 持续直到全部 `completion_confirmed` 或 paused
 7. **status.json 每步骤后由 CLI 原子写盘** — resume-safe
 8. **STATUS 枚举受限** — 仅 `DONE | DONE_WITH_CONCERNS | NEEDS_RETRY | BLOCKED`；`NEEDS_CONTEXT` 已废除
+9. **CLI 输出禁止截断** — `maestro ralph next` 的 stdout 包含完整 skill prompt，必须全量捕获。**严禁** `| head`、`| tail`、`2>&1 | head -N` 等任何截断管道。Bash timeout 可加长但不可截断输出
 </invariants>
 
 <state_machine>
@@ -153,7 +154,7 @@ Write enriched args back to status.json.
 
 ### A_EXEC_STEP
 
-1. **Load** — `run_command("maestro ralph next")`
+1. **Load** — `run_command("maestro ralph next --session <session_id>")` — **必须全量捕获 stdout，严禁 `| head`/`| tail` 等截断管道**（stdout 含完整 skill prompt，截断会导致执行内容不完整）
    - 退出码 0 → 按 stdout 内联执行
    - 退出码 2 → 交给 S_LOCATE
    - 退出码 3 → active_step_index 已被占用

package/.agy/skills/maestro-swarm-workflow/SKILL.md

@@ -20,16 +20,16 @@ swarm-workflow provides parallel compute bursts within individual steps.
 
 Scripts: `~/.maestro/workflows/swarm/wf-*.js`
 
-| Script | Accelerates | Pattern |
-|--------|-------------|---------|
-| `wf-analyze` | maestro-analyze | cli-explore-agent → 6-dimension parallel scoring → synthesis |
-| `wf-brainstorm` | maestro-brainstorm | multi-role parallel analysis → cross-role-reviewer → guidance |
-| `wf-review` | quality-review | 6-dimension workflow-reviewer → adversarial verify → report |
-| `wf-verify` | maestro-verify | 3-layer workflow-verifier + convergence + antipattern → aggregate |
-| `wf-grill` | maestro-grill | cli-explore-agent → parallel branch stress-testing → contradiction synthesis |
-| `wf-plan` | maestro-plan | parallel context exploration → workflow-planner → plan-checker |
-| `wf-execute` | maestro-execute | wave-based parallel workflow-executor (worktree isolation) |
-| `wf-milestone-audit` | maestro-milestone-audit | parallel coverage + execution + integration-checker |
+| Script | Accelerates | Adversarial Pattern |
+|--------|-------------|---------------------|
+| `wf-analyze` | maestro-analyze | explore → 6-dim scoring → **skeptic cross-verify** → **3-way advocacy (go/no-go/conditional) + referee** |
+| `wf-brainstorm` | maestro-brainstorm | multi-role analysis → **3-specialist cross-review** → **3-proposal competition** → **arbitrator** |
+| `wf-review` | quality-review | 6-dim scan → **3-vote adversarial verify (prosecutor/defense/judge)** → **3-perspective report + arbitrated verdict** |
+| `wf-verify` | maestro-verify | 3-layer + antipattern + convergence → **prosecutor vs defender debate** → **judge verdict** |
+| `wf-grill` | maestro-grill | explore → parallel branch stress → **meta-skeptic challenge** → **3-vote verdict (optimist/pessimist/realist)** |
+| `wf-plan` | maestro-plan | parallel context → **3-strategy competing proposals** → **judge panel scoring** → **3-critic adversarial check** |
+| `wf-execute` | maestro-execute | wave-based parallel execution → **adversarial convergence spot-check** → **3-vote status determination** |
+| `wf-milestone-audit` | maestro-milestone-audit | parallel 3-dim audit → **adversarial dimension challenge** → **3-vote verdict (strict/lenient/objective)** |
 
 Integration modes:
 - **Standalone**: `/maestro-swarm-workflow "analyze auth module"` — direct invocation
@@ -174,11 +174,15 @@ Intent-to-script routing（按关键词匹配，--script 优先级最高）：
 
 Workflow 返回 JSON 后：
 
-1. **摘要输出**：按脚本类型格式化关键指标
-   - analyze: overall_score, scope_verdict, go_no_go, critical findings count
-   - brainstorm: role count, conflict/synergy count, top guidance items
-   - review: verdict (APPROVE/REQUEST_CHANGES/BLOCK), confirmed vs false-positive count
-   - verify: overall_passed, confidence, gap count, antipattern blocker count
+1. **摘要输出**：按脚本类型格式化关键指标（含对抗决策结果）
+   - analyze: overall_score, scope_verdict, adversarial_outcome (go/no-go/conditional advocacy + referee), scores_challenged count
+   - brainstorm: role count, conflict/synergy count, 3-proposal competition result, arbitration notes
+   - review: verdict (APPROVE/REQUEST_CHANGES/BLOCK), 3-vote tally, confirmed vs false-positive count, adversarial_verdict
+   - verify: overall_status, prosecutor vs defender confidence, adversarial_outcome, gap count
+   - grill: overall_verdict, meta-skeptic quality rating, 3-vote verdict tally, overblown findings count
+   - plan: selected_strategy (breadth/depth/risk), judge panel scores, 3-critic adversarial check verdict
+   - execute: 3-vote status (DONE/DONE_WITH_CONCERNS/NEEDS_RETRY), convergence trust %, discrepancy count
+   - milestone-audit: 3-vote verdict, dimensions_overturned count, next_step
 
 2. **Artifact 写入**（可选）：
    - 若当前在 ralph session 中（检测 `.workflow/.maestro/ralph-*/status.json` 状态为 running）：
@@ -186,10 +190,14 @@ Workflow 返回 JSON 后：
    - 否则写入 `.workflow/scratch/{YYYYMMDD}-swarm-{script}-{slug}/results.json`
 
 3. **Ralph 兼容产出**：
-   - analyze → `analysis.md` + `context.md`（decisions）+ `conclusions.json`
-   - brainstorm → `guidance-specification.md`
-   - review → `review.json`
-   - verify → `verification.json`
+   - analyze → `analysis.md` + `context.md`（decisions）+ `conclusions.json` + `adversarial-debate.json`
+   - brainstorm → `guidance-specification.md` + `proposals-competition.json`
+   - review → `review.json`（含 adversarial_verdict + 3-vote tally）
+   - verify → `verification.json`（含 adversarial_outcome: prosecutor/defender debate）
+   - grill → `grill-results.json`（含 meta-challenge + 3-vote verdict）
+   - plan → `plan.json`（含 competition scores + critic feedback）
+   - execute → `execution-report.json`（含 convergence_checks + 3-vote status）
+   - milestone-audit → `audit-report.json`（含 dimension challenges + 3-vote verdict）
 
 4. **RunId 提示**：显示 `Resume: /maestro-swarm-workflow --resume {runId}` 用于增量重跑