Strip narrative comments and module/helper docstrings

Five rounds of sweep across the tree. Net ~544 lines removed.

Removed:
- Section-divider banners and one-line section labels (# Display
  utilities, # ----- list_requests -----, # CVSS breakdown, etc.).
- Module-level prose docstrings on internal modules. Kept one-line
  summaries; trimmed multi-paragraph narration about SDK/Strix
  responsibility splits, cache strategies, three-source precedence.
- Internal-helper docstrings that just restate the function name —
  caido_api helpers (caido_url, get_client, view_request, etc.),
  settings-class one-liners (LLMSettings, RuntimeSettings, ...),
  UI helper docstrings.
- Args/Returns blocks on non-LLM-facing internal helpers
  (build_strix_agent, render_system_prompt, create_or_reuse,
  bootstrap_caido) — kept only the genuinely non-obvious params.
- Internal-history phrasing — "Mirrors main-branch shape",
  "pre-SDK harness", "previous lookup matched no attribute".
- Narrative comments inside function bodies that explained what the
  next line does, design rationale obvious from the surrounding code,
  or "we used to..." asides.
- Trailing periods on every error-string literal across the tool tree.
- Duplicated roundtripTime quirk comment (kept the LLM-facing copy in
  tools/proxy/tools.py).

Kept (every one names an upstream bug, vendored-code provenance, or
non-obvious data quirk):
- core/runner.py: SDK replay-with-empty-initial-input + on_agent_end
  lifecycle gap.
- runtime/docker_client.py: VERBATIM COPY block of the upstream
  _create_container body, pinned to SDK v0.14.6.
- runtime/session_manager.py: NO_PROXY for agent-browser CDP loopback.
- tools/proxy/caido_api.py: generated-pydantic Request.raw quirk,
  replay double-history pitfall.
- tools/proxy/tools.py: Caido roundtripTime=0 quirk for proxy
  captures.

strix/core/agents.py

@@ -1,10 +1,4 @@
-"""SDK-native state for Strix's addressable agent graph.
-
-The Agents SDK owns model/tool execution and per-agent conversation
-history. Strix owns only product semantics the SDK does not provide:
-agent ids, the parent/child graph, wake/stop signals, TUI-visible
-status, and process-resume metadata.
-"""
+"""SDK-native state for Strix's addressable agent graph."""
 
 from __future__ import annotations
 

strix/core/execution.py

@@ -189,10 +189,7 @@ async def respawn_subagents(
     event_sink: StreamEventSink | None = None,
     hooks: RunHooks[dict[str, Any]] | None = None,
 ) -> None:
-    """Re-spawn subagent runners from a restored coordinator snapshot."""
     async with coordinator._lock:
-        # Snapshot the iteration view first so we can mutate via coordinator
-        # below without "dict changed during iteration" trouble.
         agents_snapshot = [
             (aid, status, dict(coordinator.metadata.get(aid, {})))
             for aid, status in coordinator.statuses.items()

strix/core/inputs.py

@@ -15,12 +15,10 @@ if TYPE_CHECKING:
     from strix.config.settings import ReasoningEffort
 
 
-# Default max_turns budget passed to the SDK runner.
 DEFAULT_MAX_TURNS = 500
 
 
 def build_root_task(scan_config: dict[str, Any]) -> str:
-    """Format the user-facing task for the root agent."""
     targets = scan_config.get("targets", []) or []
     diff_scope = scan_config.get("diff_scope") or {}
     user_instructions = scan_config.get("user_instructions", "") or ""
@@ -81,7 +79,6 @@ def build_root_task(scan_config: dict[str, Any]) -> str:
 
 
 def build_scope_context(scan_config: dict[str, Any]) -> dict[str, Any]:
-    """Build the system_prompt_context block consumed by the prompt template."""
     authorized: list[dict[str, str]] = []
     value_keys = {
         "repository": "target_repo",

strix/core/runner.py

@@ -1,9 +1,4 @@
-"""Top-level Strix scan runner.
-
-The SDK owns model/tool execution and per-agent sessions. This module owns
-Strix-specific scan setup, child-agent startup, resume, and the small wake loop
-needed to keep every agent addressable after its SDK run parks.
-"""
+"""Top-level Strix scan runner."""
 
 from __future__ import annotations
 
@@ -72,8 +67,6 @@ async def run_strix_scan(
     if scan_id is None:
         scan_id = f"scan-{uuid.uuid4().hex[:8]}"
 
-    # Resolve run_dir before any heavy bring-up so the log file captures
-    # everything from sandbox start onwards.
     run_dir = run_dir_for(scan_id)
     run_dir.mkdir(parents=True, exist_ok=True)
     state_dir = runtime_state_dir(run_dir)
@@ -105,15 +98,10 @@ async def run_strix_scan(
     logger.info("LLM model resolved: %s", resolved_model)
     chat_completions_tools = uses_chat_completions_tool_schema(resolved_model, settings)
 
-    # Caller may pre-create the coordinator so it can route stop/chat
-    # commands while the scan loop runs in another thread.
     if coordinator is None:
         coordinator = AgentCoordinator()
     coordinator.set_snapshot_path(agents_path)
 
-    # Wire the per-agent todo store to ``{run_dir}/.state/todos.json`` (mirrored
-    # on every CRUD) and reload any prior todos so respawned subagents
-    # find their lists intact. Same for the shared notes store.
     from strix.tools.notes.tools import hydrate_notes_from_disk
     from strix.tools.todo.tools import hydrate_todos_from_disk
 
@@ -228,8 +216,6 @@ async def run_strix_scan(
             "spawn_child_agent": spawn_child_agent,
         }
 
-        # All agents share one SQLite database; SDK session_id separates
-        # each agent's conversation inside that database.
         root_session = open_agent_session(root_id, agents_db)
         sessions_to_close.append(root_session)
         await coordinator.attach_runtime(root_id, session=root_session)
@@ -291,16 +277,8 @@ async def run_strix_scan(
         )
     except BaseException:
         logger.exception("Strix scan %s failed", scan_id)
-        # Cancel any descendant tasks the root spawned before unwinding.
-        # cancel_descendants is idempotent and handles the empty-tree case.
         if root_id is not None:
             await coordinator.cancel_descendants(root_id)
-            # The SDK's on_agent_end hook only fires after a successful
-            # ``Runner.run_streamed`` reaches the agent's first turn. A
-            # failure earlier (e.g., model-provider routing, sandbox
-            # bring-up) leaves the root stuck at status="running" — the
-            # TUI keeps animating "Initializing" forever. Finalize it
-            # here so the coordinator reflects reality.
             with contextlib.suppress(Exception):
                 await coordinator.set_status(root_id, "failed")
         raise