@researai/deepscientist 1.5.11 → 1.5.12

package/src/deepscientist/runners/codex.py

@@ -19,6 +19,11 @@ from ..shared import append_jsonl, ensure_dir, generate_id, read_yaml, resolve_r
 from ..web_search import extract_web_search_payload
 from .base import RunRequest, RunResult
 
+_TOOL_EVENT_ARGS_TEXT_LIMIT = 8_000
+_TOOL_EVENT_OUTPUT_TEXT_LIMIT = 16_000
+_MAX_QUEST_EVENT_JSON_BYTES = 2_000_000
+_OVERSIZED_EVENT_PREVIEW_TEXT_LIMIT = 12_000
+
 
 def _compact_text(value: object, *, limit: int = 1200) -> str:
     if value is None:
@@ -35,15 +40,94 @@ def _compact_text(value: object, *, limit: int = 1200) -> str:
     return text[: limit - 1].rstrip() + "…"
 
 
-def _structured_text(value: object) -> str:
+def _truncate_leaf_text(text: str, *, limit: int) -> str:
+    if limit <= 0 or len(text) <= limit:
+        return text
+    head = max(int(limit * 0.7), 256)
+    tail = max(limit - head - 64, 128)
+    omitted = max(len(text) - head - tail, 0)
+    return f"{text[:head].rstrip()}\n...[truncated {omitted} chars]...\n{text[-tail:].lstrip()}"
+
+
+def _truncate_structured_value(value: object, *, string_limit: int) -> object:
+    if isinstance(value, str):
+        return _truncate_leaf_text(value.strip(), limit=string_limit)
+    if isinstance(value, list):
+        return [_truncate_structured_value(item, string_limit=string_limit) for item in value[:200]]
+    if isinstance(value, dict):
+        truncated: dict[object, object] = {}
+        for index, (key, item) in enumerate(value.items()):
+            if index >= 200:
+                truncated["__truncated__"] = f"truncated remaining {len(value) - 200} item(s)"
+                break
+            truncated[key] = _truncate_structured_value(item, string_limit=string_limit)
+        return truncated
+    return value
+
+
+def _structured_text(value: object, *, limit: int | None = None) -> str:
     if value is None:
         return ""
     if isinstance(value, str):
-        return value.strip()
+        return _truncate_leaf_text(value.strip(), limit=limit or len(value))
+    normalized_value = _truncate_structured_value(value, string_limit=max(limit or _TOOL_EVENT_OUTPUT_TEXT_LIMIT, 512))
     try:
-        return json.dumps(value, ensure_ascii=False, indent=2)
+        return json.dumps(normalized_value, ensure_ascii=False, indent=2)
     except TypeError:
-        return str(value)
+        return _truncate_leaf_text(str(value), limit=limit or _TOOL_EVENT_OUTPUT_TEXT_LIMIT)
+
+
+def _encoded_json_size(value: object) -> int:
+    try:
+        return len(json.dumps(value, ensure_ascii=False).encode("utf-8"))
+    except Exception:
+        return len(str(value).encode("utf-8", errors="ignore"))
+
+
+def _compact_tool_event_payload(payload: dict[str, Any]) -> dict[str, Any]:
+    if _encoded_json_size(payload) <= _MAX_QUEST_EVENT_JSON_BYTES:
+        return payload
+
+    compacted = dict(payload)
+    output_text = str(compacted.get("output") or "")
+    if output_text:
+        compacted["output_bytes"] = len(output_text.encode("utf-8", errors="ignore"))
+        compacted["output"] = _truncate_leaf_text(
+            output_text,
+            limit=_OVERSIZED_EVENT_PREVIEW_TEXT_LIMIT,
+        )
+        compacted["output_truncated"] = True
+    args_text = str(compacted.get("args") or "")
+    if args_text and _encoded_json_size(compacted) > _MAX_QUEST_EVENT_JSON_BYTES:
+        compacted["args"] = _truncate_leaf_text(args_text, limit=4_000)
+        compacted["args_truncated"] = True
+    if _encoded_json_size(compacted) > _MAX_QUEST_EVENT_JSON_BYTES:
+        metadata = compacted.get("metadata")
+        if isinstance(metadata, dict):
+            allowed_keys = {
+                "mcp_server",
+                "mcp_tool",
+                "bash_id",
+                "status",
+                "command",
+                "workdir",
+                "cwd",
+                "started_at",
+                "finished_at",
+                "exit_code",
+                "stop_reason",
+                "log_path",
+            }
+            compacted["metadata"] = {
+                key: metadata.get(key)
+                for key in allowed_keys
+                if key in metadata
+            }
+            compacted["metadata_truncated"] = True
+    if _encoded_json_size(compacted) > _MAX_QUEST_EVENT_JSON_BYTES:
+        compacted["output"] = _compact_text(compacted.get("output"), limit=2_000)
+        compacted["output_truncated"] = True
+    return compacted
 
 
 def _iter_event_texts(event: dict[str, Any]) -> list[str]:
@@ -209,7 +293,7 @@ def _tool_args(event: dict[str, Any], item: dict[str, Any]) -> str:
             item.get("input"),
             event.get("input"),
         ):
-            text = _structured_text(value)
+            text = _structured_text(value, limit=_TOOL_EVENT_ARGS_TEXT_LIMIT)
             if text:
                 return text
         return ""
@@ -243,7 +327,7 @@ def _tool_output(event: dict[str, Any], item: dict[str, Any]) -> str:
             item.get("aggregated_output"),
             event.get("aggregated_output"),
         ):
-            text = _structured_text(value)
+            text = _structured_text(value, limit=_TOOL_EVENT_OUTPUT_TEXT_LIMIT)
             if text:
                 return text
         return ""
@@ -361,7 +445,7 @@ def _tool_event(
                 "raw_event_type": event_type,
                 "created_at": created_at,
             }
-        return {
+        return _compact_tool_event_payload({
             "event_id": generate_id("evt"),
             "type": "runner.tool_result",
             "quest_id": quest_id,
@@ -375,7 +459,7 @@ def _tool_event(
             "output": _tool_output(event, item),
             "raw_event_type": event_type,
             "created_at": created_at,
-        }
+        })
 
     if item_type == "web_search":
         tool_call_id = _tool_call_id(event, item)
@@ -399,7 +483,7 @@ def _tool_event(
                 "raw_event_type": event_type,
                 "created_at": created_at,
             }
-        return {
+        return _compact_tool_event_payload({
             "event_id": generate_id("evt"),
             "type": "runner.tool_result",
             "quest_id": quest_id,
@@ -414,13 +498,13 @@ def _tool_event(
             "metadata": metadata,
             "raw_event_type": event_type,
             "created_at": created_at,
-        }
+        })
 
     if item_type == "file_change":
         tool_call_id = _tool_call_id(event, item)
         tool_name = "file_change"
         known_tool_names[tool_call_id] = tool_name
-        return {
+        return _compact_tool_event_payload({
             "event_id": generate_id("evt"),
             "type": "runner.tool_result",
             "quest_id": quest_id,
@@ -433,7 +517,7 @@ def _tool_event(
             "output": _tool_output(event, item),
             "raw_event_type": event_type,
             "created_at": created_at,
-        }
+        })
 
     if item_type == "mcp_tool_call":
         tool_call_id = _tool_call_id(event, item)
@@ -466,7 +550,7 @@ def _tool_event(
                 "raw_event_type": event_type,
                 "created_at": created_at,
             }
-        return {
+        return _compact_tool_event_payload({
             "event_id": generate_id("evt"),
             "type": "runner.tool_result",
             "quest_id": quest_id,
@@ -483,7 +567,7 @@ def _tool_event(
             "metadata": metadata,
             "raw_event_type": event_type,
             "created_at": created_at,
-        }
+        })
 
     if item_type in {"function_call", "custom_tool_call", "tool_call"} or "function_call" in event_type or "tool_call" in event_type:
         tool_call_id = _tool_call_id(event, item)
@@ -507,7 +591,7 @@ def _tool_event(
     if item_type in {"function_call_output", "custom_tool_call_output", "tool_result", "tool_call_output"} or "function_call_output" in event_type or "tool_result" in event_type:
         tool_call_id = _tool_call_id(event, item)
         tool_name = known_tool_names.get(tool_call_id) or _tool_name(event, item)
-        return {
+        return _compact_tool_event_payload({
             "event_id": generate_id("evt"),
             "type": "runner.tool_result",
             "quest_id": quest_id,
@@ -521,7 +605,7 @@ def _tool_event(
             "output": _tool_output(event, item),
             "raw_event_type": event_type,
             "created_at": created_at,
-        }
+        })
 
     return None
 
@@ -582,6 +666,12 @@ class CodexRunner:
         )
 
         env = dict(**os.environ)
+        runner_env = runner_config.get("env") if isinstance(runner_config.get("env"), dict) else {}
+        for key, value in runner_env.items():
+            env_key = str(key or "").strip()
+            if not env_key or value is None:
+                continue
+            env[env_key] = str(value)
         env["CODEX_HOME"] = str(codex_home)
         env["DEEPSCIENTIST_HOME"] = str(self.home)
         env["DS_HOME"] = str(self.home)
@@ -809,16 +899,23 @@ class CodexRunner:
         workspace_root = request.worktree_root or request.quest_root
         resolved_binary = resolve_runner_binary(self.binary, runner_name="codex")
         resolved_runner_config = runner_config if isinstance(runner_config, dict) else self._load_runner_config()
+        profile = str(resolved_runner_config.get("profile") or "").strip()
         normalized_model = str(request.model or "").strip()
         command = [
             resolved_binary or self.binary,
             "--search",
-            "exec",
-            "--json",
-            "--cd",
-            str(workspace_root),
-            "--skip-git-repo-check",
         ]
+        if profile:
+            command.extend(["--profile", profile])
+        command.extend(
+            [
+                "exec",
+                "--json",
+                "--cd",
+                str(workspace_root),
+                "--skip-git-repo-check",
+            ]
+        )
         if normalized_model.lower() not in {"", "inherit", "default", "codex-default"}:
             command.extend(["--model", normalized_model])
         if request.approval_policy:
@@ -846,7 +943,9 @@ class CodexRunner:
         runner_config: dict[str, Any] | None = None,
     ) -> Path:
         target = ensure_dir(workspace_root / ".codex")
-        source = Path(os.environ.get("CODEX_HOME", str(Path.home() / ".codex"))).expanduser()
+        resolved_runner_config = runner_config if isinstance(runner_config, dict) else self._load_runner_config()
+        configured_home = str(resolved_runner_config.get("config_dir") or os.environ.get("CODEX_HOME") or str(Path.home() / ".codex"))
+        source = Path(configured_home).expanduser()
         for filename in ("config.toml", "auth.json"):
             source_path = source / filename
             target_path = target / filename

package/src/deepscientist/runners/runtime_overrides.py

@@ -20,6 +20,8 @@ def _as_bool_env(name: str) -> bool:
 def codex_runtime_overrides() -> dict[str, str]:
     approval_policy = _as_text(os.environ.get("DEEPSCIENTIST_CODEX_APPROVAL_POLICY"))
     sandbox_mode = _as_text(os.environ.get("DEEPSCIENTIST_CODEX_SANDBOX_MODE"))
+    profile = _as_text(os.environ.get("DEEPSCIENTIST_CODEX_PROFILE"))
+    model = _as_text(os.environ.get("DEEPSCIENTIST_CODEX_MODEL"))
 
     if _as_bool_env("DEEPSCIENTIST_CODEX_YOLO"):
         approval_policy = approval_policy or "never"
@@ -30,6 +32,10 @@ def codex_runtime_overrides() -> dict[str, str]:
         overrides["approval_policy"] = approval_policy
     if sandbox_mode:
         overrides["sandbox_mode"] = sandbox_mode
+    if profile:
+        overrides["profile"] = profile
+    if model:
+        overrides["model"] = model
     return overrides
 
 

package/src/deepscientist/shared.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+from collections import deque
 import hashlib
 import json
 import os
@@ -9,7 +10,7 @@ import subprocess
 import sys
 from datetime import UTC, datetime
 from pathlib import Path
-from typing import Any
+from typing import Any, Iterator
 from uuid import uuid4
 
 try:
@@ -90,21 +91,39 @@ def append_jsonl(path: Path, payload: dict[str, Any]) -> None:
         handle.write(json.dumps(payload, ensure_ascii=False) + "\n")
 
 
-def read_jsonl(path: Path) -> list[dict[str, Any]]:
+def iter_jsonl(path: Path | str) -> Iterator[dict[str, Any]]:
+    path = Path(path)
     if not path.exists():
+        return
+    with path.open("r", encoding="utf-8") as handle:
+        for raw_line in handle:
+            line = raw_line.strip()
+            if not line:
+                continue
+            try:
+                payload = json.loads(line)
+            except json.JSONDecodeError:
+                continue
+            if isinstance(payload, dict):
+                yield payload
+
+
+def read_jsonl(path: Path) -> list[dict[str, Any]]:
+    return list(iter_jsonl(path))
+
+
+def count_jsonl(path: Path | str) -> int:
+    return sum(1 for _ in iter_jsonl(path))
+
+
+def read_jsonl_tail(path: Path | str, limit: int) -> list[dict[str, Any]]:
+    normalized_limit = max(int(limit or 0), 0)
+    if normalized_limit <= 0:
         return []
-    items: list[dict[str, Any]] = []
-    for line in path.read_text(encoding="utf-8").splitlines():
-        line = line.strip()
-        if not line:
-            continue
-        try:
-            payload = json.loads(line)
-        except json.JSONDecodeError:
-            continue
-        if isinstance(payload, dict):
-            items.append(payload)
-    return items
+    items: deque[dict[str, Any]] = deque(maxlen=normalized_limit)
+    for payload in iter_jsonl(path):
+        items.append(payload)
+    return list(items)
 
 
 def read_yaml(path: Path, default: Any = None) -> Any:

package/src/prompts/connectors/qq.md

@@ -10,7 +10,8 @@
 - qq_summary_first_rule: start with the conclusion the user cares about, then what it means, then the next action
 - qq_progress_shape_rule: make the current task, the main difficulty or latest real progress, and the next concrete measure explicit whenever possible
 - qq_eta_rule: for baseline reproduction, main experiments, analysis experiments, and other important long-running research phases, include a rough ETA for the next meaningful result or the next update; if uncertain, say that and still give the next check-in window
-- qq_tool_call_keepalive_rule: for ordinary active work, prefer one concise QQ progress update after roughly 10 tool calls when there is already a human-meaningful delta, and do not let work drift beyond roughly 20 tool calls or about 15 minutes without a user-visible checkpoint
+- qq_tool_call_keepalive_rule: for ordinary active work, prefer one concise QQ progress update after roughly 6 tool calls when there is already a human-meaningful delta, and do not let work drift beyond roughly 12 tool calls or about 8 minutes without a user-visible checkpoint
+- qq_read_plan_keepalive_rule: if the active work is still mostly reading, comparison, or planning, do not wait too long for a "big result"; send a short QQ-facing checkpoint after about 5 consecutive tool calls if the user would otherwise see silence
 - qq_internal_detail_rule: omit worker names, heartbeat timestamps, retry counters, pending/running/completed counts, file names, and monitor-window narration unless the user asked for them or the detail changes the recommended action
 - qq_translation_rule: convert internal execution and file-management work into user value, such as saying the baseline record is now organized for easier later comparison instead of listing touched files
 - qq_preflight_rule: before sending a QQ progress update, rewrite it if it still sounds like a monitoring log, execution diary, or file inventory

package/src/prompts/connectors/weixin.md

@@ -10,7 +10,8 @@
 - weixin_summary_first_rule: start with the user-facing conclusion, then what it means, then the next action
 - weixin_progress_shape_rule: make the current task, the main difficulty or latest real progress, and the next concrete measure explicit whenever possible
 - weixin_eta_rule: for important long-running phases such as baseline reproduction, main experiments, analysis, or paper packaging, include a rough ETA or next check-in window when you can
-- weixin_tool_call_keepalive_rule: for ordinary active work, prefer one concise Weixin progress update after roughly 10 tool calls when there is already a human-meaningful delta, and do not let work drift beyond roughly 20 tool calls or about 15 minutes without a user-visible checkpoint
+- weixin_tool_call_keepalive_rule: for ordinary active work, prefer one concise Weixin progress update after roughly 6 tool calls when there is already a human-meaningful delta, and do not let work drift beyond roughly 12 tool calls or about 8 minutes without a user-visible checkpoint
+- weixin_read_plan_keepalive_rule: if the active work is still mostly reading, comparison, or planning, do not wait too long for a "big result"; send a short Weixin-facing checkpoint after about 5 consecutive tool calls if the user would otherwise see silence
 - weixin_internal_detail_rule: omit worker names, retry counters, pending/running/completed counts, low-level file listings, and monitor-window narration unless the user explicitly asked for them or they change the recommended action
 - weixin_translation_rule: translate internal execution and file-management work into user value instead of narrating tool or filesystem churn
 - weixin_preflight_rule: before sending a Weixin-facing progress update, rewrite it if it still reads like a monitor log, execution diary, or file inventory

package/src/prompts/contracts/shared_interaction.md

@@ -7,7 +7,10 @@ This shared contract is injected once per turn and applies across the stage and
 - Treat `artifact.interact(...)` as the main long-lived communication thread across TUI, web, and bound connectors.
 - If `artifact.interact(...)` returns queued user requirements, treat them as the highest-priority user instruction bundle before continuing the current stage or companion-skill task.
 - Immediately follow any non-empty mailbox poll with another `artifact.interact(...)` update that confirms receipt; if the request is directly answerable, answer there, otherwise say the current subtask is paused, give a short plan plus nearest report-back point, and handle that request first.
-- Emit `artifact.interact(kind='progress', reply_mode='threaded', ...)` when there is real user-visible progress: a meaningful checkpoint, route-shaping update, or a concise keepalive once active work has crossed roughly 10 tool calls with a human-meaningful delta. Do not let ordinary active work drift beyond roughly 20 tool calls or about 15 minutes without a user-visible update.
+- Stage-kickoff rule: after entering any stage or companion skill, send one `artifact.interact(kind='progress', reply_mode='threaded', ...)` update within the first 3 tool calls of substantial work.
+- Reading/planning keepalive rule: if you spend 5 consecutive tool calls on reading, searching, comparison, or planning without a user-visible update, send one concise checkpoint even if the route is not finalized yet.
+- Subtask-boundary rule: send a user-visible update whenever the active subtask changes materially, especially across intake -> audit, audit -> experiment planning, experiment planning -> run launch, run result -> drafting, or drafting -> review/rebuttal.
+- Emit `artifact.interact(kind='progress', reply_mode='threaded', ...)` when there is real user-visible progress: a meaningful checkpoint, route-shaping update, or a concise keepalive once active work has crossed roughly 6 tool calls with a human-meaningful delta. Do not let ordinary active work drift beyond roughly 12 tool calls or about 8 minutes without a user-visible update.
 - Keep progress updates chat-like and easy to understand: say what changed, what it means, and what happens next.
 - Default to plain-language summaries. Do not mention file paths, artifact ids, branch/worktree ids, session ids, raw commands, or raw logs unless the user asks or needs them to act.
 - Use `reply_mode='blocking'` only for real user decisions that cannot be resolved from local evidence.

package/src/prompts/system.md

@@ -53,7 +53,7 @@ Your job is to keep a research quest moving forward in a durable, auditable, evi
   - for ordinary progress replies, usually stay within 2 to 4 short sentences or 3 short bullets at most
   - start with the conclusion the user cares about, then what it means, then the next action
   - for baseline reproduction, main experiments, analysis experiments, and similar long-running research phases, also tell the user roughly how long until the next meaningful result, next step, or next update
-  - for ordinary active multi-step work, prefer a concise update once active work has crossed about 10 tool calls and there is already a human-meaningful delta, and do not disappear for more than about 20 tool calls or about 15 minutes of active foreground work without a user-visible update unless a real milestone is imminent
+  - for ordinary active multi-step work, prefer a concise update once active work has crossed about 6 tool calls and there is already a human-meaningful delta, and do not disappear for more than about 12 tool calls or about 8 minutes of active foreground work without a user-visible update unless a real milestone is imminent
   - do not spam internal tool chatter, raw diffs, or every small checkpoint
   - do not proactively enumerate file paths, file inventories, or low-level file details unless the user explicitly asks
   - do not proactively expose worker names, heartbeat timestamps, retry counters, pending/running/completed counts, or monitor-window narration unless that detail changes the recommended action or is required for honesty about risk
@@ -203,7 +203,7 @@ When you send user-facing updates (especially via `artifact.interact(...)`), wri
   - what task you are currently working on
   - what the main difficulty, risk, or latest real progress is
   - what concrete next step or mitigation you will take
-- for ordinary active multi-step work, if no natural milestone arrives, prefer a short progress update once active work has crossed about 10 tool calls and there is already a human-meaningful delta, and do not drift beyond about 20 tool calls or about 15 minutes of active foreground work without any user-visible checkpoint
+- for ordinary active multi-step work, if no natural milestone arrives, prefer a short progress update once active work has crossed about 6 tool calls and there is already a human-meaningful delta, and do not drift beyond about 12 tool calls or about 8 minutes of active foreground work without any user-visible checkpoint
 - for baseline reproduction, main experiments, analysis experiments, and similar long-running phases, also make the timing expectation explicit:
   - roughly how long until the next meaningful result, next milestone, or next update, usually within a 10 to 30 minute window
   - if runtime is uncertain, say that directly and give the next check-in window instead of pretending to know an exact ETA
@@ -463,9 +463,12 @@ Each milestone update should usually state:
 Cadence defaults for ordinary active work:
 
 - treat `artifact.interact(...)` as the default user-visible heartbeat rather than an optional extra
-- soft trigger: after about 10 tool calls, if there is already a human-meaningful delta, send `artifact.interact(kind='progress', reply_mode='threaded', ...)`
-- hard trigger: do not exceed about 20 tool calls without a user-visible `artifact.interact(...)` update during active foreground work
-- time trigger: do not exceed about 15 minutes of active foreground work without a user-visible update, even if the tool-call count stayed low
+- stage-kickoff trigger: after entering any stage or companion skill, send one `artifact.interact(kind='progress', reply_mode='threaded', ...)` update within the first 3 tool calls of substantial work
+- reading/planning trigger: if you spend about 5 consecutive tool calls on reading, searching, comparison, or planning without a user-visible update, send one concise checkpoint even if the route is not finalized yet
+- boundary trigger: send a user-visible update whenever the active subtask changes materially, especially across intake -> audit, audit -> experiment planning, experiment planning -> run launch, run result -> drafting, or drafting -> review/rebuttal
+- soft trigger: after about 6 tool calls, if there is already a human-meaningful delta, send `artifact.interact(kind='progress', reply_mode='threaded', ...)`
+- hard trigger: do not exceed about 12 tool calls without a user-visible `artifact.interact(...)` update during active foreground work
+- time trigger: do not exceed about 8 minutes of active foreground work without a user-visible update, even if the tool-call count stayed low
 - immediate trigger: send a user-visible update as soon as a real blocker, recovery, route change, branch/worktree switch, baseline gate change, selected idea, recorded main experiment, or user-priority interruption becomes clear
 - de-duplication rule: do not send another ordinary progress update within about 2 additional tool calls or about 90 seconds unless a real milestone, blocker, route change, or new user message makes that extra update genuinely useful
 - keep ordinary subtask completions short; reserve richer milestone reports for stage-significant deliverables and route-changing checkpoints instead of narrating every small setup step
@@ -1080,9 +1083,10 @@ For `artifact.interact(...)` specifically:
   - raw logs
   - internal tool names
 - mention those details only if the user asked for them or needs them to act on the message
-- during active work, emit `artifact.interact(kind='progress', ...)` at real human-meaningful checkpoints; if no natural checkpoint appears, prefer sending one once active work has crossed about 10 tool calls and there is already a human-meaningful delta, and do not drift beyond about 20 tool calls or about 15 minutes of active foreground work without a user-visible update
+- during active work, emit `artifact.interact(kind='progress', ...)` at real human-meaningful checkpoints; if no natural checkpoint appears, prefer sending one once active work has crossed about 6 tool calls and there is already a human-meaningful delta, and do not drift beyond about 12 tool calls or about 8 minutes of active foreground work without a user-visible update
 - during long active execution, after the first meaningful signal from long-running work, keep the user informed and never let active user-relevant work go more than 30 minutes without a real progress inspection and, if still running, a user-visible keepalive
-- do not send another ordinary progress update within about 2 additional tool calls or about 90 seconds unless a milestone, blocker, route change, or new user message makes it genuinely useful
+- if the active work is still mostly reading, comparison, synthesis, or planning, do not hide behind "no result yet"; send a short user-visible checkpoint after about 5 consecutive tool calls if the user would otherwise see silence
+- do not send another ordinary progress update within about 2 additional tool calls or about 60 seconds unless a milestone, blocker, route change, or new user message makes it genuinely useful
 - each ordinary progress update should usually answer only:
   - what changed
   - what it means now
@@ -1321,7 +1325,7 @@ If the field is absent, default to `freeform`.
 When `launch_mode = custom`:
 
 - do not force the quest back into the canonical full-research path if the custom brief is narrower
-- treat `entry_state_summary`, `review_summary`, and `custom_brief` as real startup context rather than decorative metadata
+- treat `entry_state_summary`, `review_summary`, `review_materials`, and `custom_brief` as real startup context rather than decorative metadata
 - if the quest clearly starts from existing baseline / result / draft state, open `intake-audit` before restarting baseline discovery or fresh experimentation
 - if the quest clearly starts from reviewer comments, a revision request, or a rebuttal packet, open `rebuttal` before ordinary `write`
 - after the custom entry skill stabilizes the route, continue through the normal stage skills as needed
@@ -1331,12 +1335,58 @@ When `custom_profile = continue_existing_state`:
 - assume the quest may already contain reusable baselines, measured results, analysis assets, or writing assets
 - audit and trust-rank those assets first instead of reflexively rerunning everything
 
+When `custom_profile = review_audit`:
+
+- assume the active contract is a substantial draft or paper package that needs an independent skeptical audit
+- open `review` before more writing or finalization
+- if the audit finds real gaps, route to the needed downstream skill instead of polishing blindly
+
+When `startup_contract.review_followup_policy = auto_execute_followups`:
+
+- after review artifacts are durable, continue automatically into the required experiments, manuscript deltas, and review-closure work
+- do not stop at the audit report if the route is already clear
+
+When `startup_contract.review_followup_policy = user_gated_followups`:
+
+- finish the review artifacts first
+- then raise one structured decision before expensive experiments or manuscript revisions continue
+
+When `startup_contract.review_followup_policy = audit_only`:
+
+- stop after the durable audit artifacts and route recommendation unless the user later asks for execution follow-up
+
 When `custom_profile = revision_rebuttal`:
 
 - assume the active contract is a paper-review workflow rather than a blank research loop
 - preserve the existing paper, results, and reviewer package as the starting state
 - route supplementary experiments through `analysis-campaign` and manuscript deltas through `write`, but let `rebuttal` orchestrate that mapping
 
+When `startup_contract.baseline_execution_policy = must_reproduce_or_verify`:
+
+- explicitly verify or recover the rebuttal-critical baseline or comparator before reviewer-linked follow-up work
+
+When `startup_contract.baseline_execution_policy = reuse_existing_only`:
+
+- trust the current confirmed baseline/results unless you find concrete inconsistency, corruption, or missing-evidence problems
+
+When `startup_contract.baseline_execution_policy = skip_unless_blocking`:
+
+- do not spend time rerunning baselines by default
+- only open `baseline` if a named review/rebuttal issue truly depends on a missing comparator or unusable prior evidence
+
+When `startup_contract.manuscript_edit_mode = latex_required`:
+
+- if manuscript revision is required, treat the provided LaTeX tree or `paper/latex/` as the writing surface
+- if LaTeX source is unavailable, do not pretend the manuscript was edited; produce LaTeX-ready replacement text and state the blocker explicitly
+
+When `startup_contract.manuscript_edit_mode = copy_ready_text`:
+
+- provide section-level copy-ready replacement text and explicit deltas when manuscript revision is required
+
+When `startup_contract.manuscript_edit_mode = none`:
+
+- revision planning artifacts are sufficient unless the user later broadens scope
+
 When `custom_profile = freeform`:
 
 - treat the custom brief as the primary scope contract
@@ -2078,7 +2128,7 @@ When summarizing long logs, campaigns, or multi-agent work:
   - the estimated next reply time (usually the next sleep interval you are about to use)
 - If the run still looks healthy but there is no human-meaningful delta yet, continue monitoring silently instead of sending a no-change keepalive just because a sleep finished.
 - For baseline reproduction, main experiments, analysis experiments, and similar user-relevant long runs, translate that monitoring ETA into user-facing language such as how long until the next meaningful result or the next expected update.
-- Outside those detached experiment waits, prefer sending a concise `artifact.interact(kind='progress', ...)` once active work has crossed about 10 tool calls and there is already a human-meaningful delta, and do not let active foreground work drift beyond about 20 tool calls or about 15 minutes without a user-visible checkpoint.
+- Outside those detached experiment waits, prefer sending a concise `artifact.interact(kind='progress', ...)` once active work has crossed about 6 tool calls and there is already a human-meaningful delta, and do not let active foreground work drift beyond about 12 tool calls or about 8 minutes without a user-visible checkpoint.
 - If you forget a bash id, do not guess. Use `bash_exec(mode='history')` or `bash_exec(mode='list')` and recover it from the reverse-chronological session list.
 - If the long-running command or wrapper code can emit structured progress markers, prefer a concise `__DS_PROGRESS__ { ... }` JSON line with fields such as:
   - `current`