fix(ui): track last_tool_call_id; paint fresh chamber on parallel-result mismatch

Closes dirge-jzj.

The UI's chamber state model was single-slot (one `tool_chamber_open:
bool` + one `last_tool_name: Option<String>`). Under parallel tool
execution — the default per `agent_loop/types.rs:402` since this
matches pi — the `execute_tool_calls_parallel` loop fires
`ToolExecutionStart` for ALL tool calls in a turn back-to-back before
any `ToolExecutionEnd`. Each new ToolCall passively closes the
prior chamber via the close at line 2137. ToolResults arrive in
completion order, which often differs from call order.

Symptoms before this fix (after 7403792's passive-close change):
- `╭─ READ ─ "/foo" ─╮ ╰────╯` empty chamber for the displaced call
- `↳ first_line` trailer painted below an unrelated chamber when the
  displaced call's result finally arrives
- Pre-7403792 the same race produced false "⚠ tool denied · aborted ·
  no result" wording instead of empty chambers; same root cause, just
  a different visible symptom.

Fix: track `last_tool_call_id: Option<String>` alongside the existing
`last_tool_name`. Set at ToolCall, cleared at ToolResult end and
chamber-close-during-result paths. At ToolResult, if the result's id
doesn't match `last_tool_call_id`, the chamber on screen belongs to
a sibling call — paint a fresh complete chamber for THIS id below
the current scroll position (close current passive, look up
name+args from `tool_calls_buf`, paint TOP, set chamber_open=true).
The existing render paths (a/b/c) then handle the body for the
newly-opened chamber.

Each parallel ToolResult lands in its own correctly-labeled frame,
rendered in completion order. The display can appear out of source
order when tools finish at different speeds — that's accurate to
what happened.

Trade-off: deny flow with parallel calls + alert response now paints
a fresh chamber containing the deny error text below the existing
`↳ denied: tool args` trailer line. Slightly redundant but
informative. Threading call_id through the ask flow would let us
suppress the redundancy; deferred to a follow-up since the
parallel-deny intersection is rare.

The id-matches case (sequential / single-tool-per-turn, which is
the common path) falls through to existing path (a) unchanged.

Tests: existing 192 ui:: tests still pass. The multi-chamber
behavior is event-stream-sequenced and not easily covered by the
current unit-test harness (which tests chamber close helpers in
isolation, not full event loops); integration test is a follow-up.

Author: Yogthos

.beads/issues.jsonl

@@ -2,7 +2,7 @@
 {"_type":"issue","id":"dirge-ojn","title":"Perm M4: flip unmatched-tool default from Allow to Ask","description":"Today src/permission/checker.rs:103 has default_action = Action::Allow. Anything not explicitly configured falls through to Allow without prompting — a security gap especially for write/edit/apply_patch which currently have NO default rules installed at all (mismatch with bash which has default_bash_rules, mcp_tool which defaults Ask since 27bd70a).\n\nFlip default to Ask. Add builtin-allow list for safe read-only tools (port maki's BUILTIN_ALLOW_RULES at permissions.rs:16-24, adapted for dirge's tool set: read/glob/grep/list_dir/list_symbols/find_definition/find_callers/find_callees/get_symbol_body/repo_overview).\n\nwrite/edit/apply_patch/bash/webfetch/websearch/task/skill/memory all become Ask by default unless explicitly allowlisted. Document the migration in README + CHANGELOG. Users with existing configs unaffected — only the no-config baseline changes.\n\nAdd --yolo CLI flag (currently only via config). Maps to set the global allow_all atomic (port maki's PermissionManager::toggle_yolo at permissions.rs:219-222).\n\nDepends on: dirge-{M2}, dirge-{M3}","status":"open","priority":1,"issue_type":"feature","owner":"yogthos@gmail.com","created_at":"2026-05-23T14:25:47Z","created_by":"Yogthos","updated_at":"2026-05-23T14:25:47Z","dependencies":[{"issue_id":"dirge-ojn","depends_on_id":"dirge-6ab","type":"blocks","created_at":"2026-05-23T10:25:53Z","created_by":"Yogthos","metadata":"{}"},{"issue_id":"dirge-ojn","depends_on_id":"dirge-cep","type":"blocks","created_at":"2026-05-23T10:25:52Z","created_by":"Yogthos","metadata":"{}"}],"dependency_count":2,"dependent_count":0,"comment_count":0}
 {"_type":"issue","id":"dirge-01s","title":"Perm M1: single chokepoint (port maki's enforce shape)","description":"Refactor dirge's 3 permission entry points (check_perm, check_perm_path, check_perm_path_resolve in src/agent/tools/mod.rs) into a single chokepoint patterned after maki's PermissionManager::enforce (maki-agent/src/permissions.rs:283). One function, takes (tool, scope), routes internally based on tool category. No behavior change for users — pure refactor with all existing callers updated. Behind-the-scenes the new fn calls the same PermissionChecker logic.\n\nRef: maki-agent/src/permissions.rs:283-350 — port the signature shape (async, returns Result\u003c(), PermissionError\u003e, takes \u0026EventSender + user_response_rx + cancel for UI integration). Adapt to dirge's existing AskSender + PermCheck.","status":"open","priority":1,"issue_type":"feature","owner":"yogthos@gmail.com","created_at":"2026-05-23T14:25:25Z","created_by":"Yogthos","updated_at":"2026-05-23T14:25:25Z","dependency_count":0,"dependent_count":2,"comment_count":0}
 {"_type":"issue","id":"dirge-cep","title":"Perm M2: unified rule schema (port maki's TOML shape via JSON)","description":"Replace dirge's per-tool field PermissionConfig (separate Option\u003cToolPerm\u003e per built-in tool) with maki's uniform shape: { tool_name: { allow: [patterns], deny: [patterns] } }. Single PermissionRule struct (maki-config/src/lib.rs:268-273). Keep existing JSON config (dirge uses JSON not TOML), but flatten the schema. Add a dual-read path so old config.json files (existing user configs in the wild) still parse — log a deprecation warning and auto-migrate on save. Migration test required.\n\nDepends on: dirge-{M1}","status":"open","priority":1,"issue_type":"feature","owner":"yogthos@gmail.com","created_at":"2026-05-23T14:25:25Z","created_by":"Yogthos","updated_at":"2026-05-23T14:25:25Z","dependencies":[{"issue_id":"dirge-cep","depends_on_id":"dirge-01s","type":"blocks","created_at":"2026-05-23T10:25:50Z","created_by":"Yogthos","metadata":"{}"}],"dependency_count":1,"dependent_count":1,"comment_count":0}
-{"_type":"issue","id":"dirge-jzj","title":"UI: per-tool-call-id chamber state (parallel tool execution)","description":"The UI tracks chamber state via a single bool tool_chamber_open + Option\u003cString\u003e last_tool_name. Under parallel tool execution (the default per agent_loop/types.rs:402), multiple ToolExecutionStart events fire before any ToolExecutionEnd. Subsequent ToolCalls close prior chambers prematurely (now via passive close after 7403792, previously with false 'denied' wording). ToolResults that arrive after a newer ToolCall's chamber opens land as out-of-place '↳ trailers' below the wrong chamber. Fix: chamber state keyed by tool_call_id (HashMap), each in-flight tool gets its own chamber frame, the abort/passive close functions take an id.","status":"open","priority":1,"issue_type":"bug","owner":"yogthos@gmail.com","created_at":"2026-05-23T14:25:11Z","created_by":"Yogthos","updated_at":"2026-05-23T14:25:11Z","dependency_count":0,"dependent_count":0,"comment_count":0}
+{"_type":"issue","id":"dirge-jzj","title":"UI: per-tool-call-id chamber state (parallel tool execution)","description":"The UI tracks chamber state via a single bool tool_chamber_open + Option\u003cString\u003e last_tool_name. Under parallel tool execution (the default per agent_loop/types.rs:402), multiple ToolExecutionStart events fire before any ToolExecutionEnd. Subsequent ToolCalls close prior chambers prematurely (now via passive close after 7403792, previously with false 'denied' wording). ToolResults that arrive after a newer ToolCall's chamber opens land as out-of-place '↳ trailers' below the wrong chamber. Fix: chamber state keyed by tool_call_id (HashMap), each in-flight tool gets its own chamber frame, the abort/passive close functions take an id.","status":"closed","priority":1,"issue_type":"bug","owner":"yogthos@gmail.com","created_at":"2026-05-23T14:25:11Z","created_by":"Yogthos","updated_at":"2026-05-23T14:35:50Z","started_at":"2026-05-23T14:27:32Z","closed_at":"2026-05-23T14:35:50Z","close_reason":"Closed","dependency_count":0,"dependent_count":0,"comment_count":0}
 {"_type":"issue","id":"dirge-5uv","title":"H-batch1-2: history up/down byte-offset panic risk on multi-byte input","description":"ui/input.rs:988-1023 — col = self.cursor - line_start is byte distance; self.cursor = (pos + col).min(target_line_end) can land mid-codepoint when previous line has multi-byte chars. Subsequent replace_range/slice panics. Convert column to char index then back to byte offset on target line.","status":"closed","priority":1,"issue_type":"bug","owner":"yogthos@gmail.com","created_at":"2026-05-23T02:34:57Z","created_by":"Yogthos","updated_at":"2026-05-23T03:09:57Z","closed_at":"2026-05-23T03:09:57Z","close_reason":"Closed","dependency_count":0,"dependent_count":0,"comment_count":0}
 {"_type":"issue","id":"dirge-le5","title":"H-batch1-1: bash inherits all env vars including API keys","description":"tools/bash.rs:208-320 no .env_clear(). Sensitive vars (OPENROUTER_API_KEY, EXA_API_KEY, PARALLEL_API_KEY, ANTHROPIC_API_KEY, etc.) flow to every bash child. Use .env_clear() + curated allowlist (PATH, HOME, USER, LANG, TERM, etc.). See pi bash-executor.ts.","status":"closed","priority":1,"issue_type":"bug","owner":"yogthos@gmail.com","created_at":"2026-05-23T02:34:57Z","created_by":"Yogthos","updated_at":"2026-05-23T03:09:57Z","closed_at":"2026-05-23T03:09:57Z","close_reason":"Closed","dependency_count":0,"dependent_count":0,"comment_count":0}
 {"_type":"issue","id":"dirge-846","title":"C8: /cd, /worktree, /wt-exit, /prompt, /regen-prompts silently drop LSP","description":"ui/slash.rs:1290, 1300, 1763 + analogues in mod.rs — every build_agent call after these commands passes None for lsp_manager. User silently loses LSP for the rest of the session after first /cd. Re-thread lsp_manager (and re-root for /cd/worktree). opencode cd handler re-roots LSP.","status":"closed","priority":1,"issue_type":"bug","assignee":"Yogthos","owner":"yogthos@gmail.com","created_at":"2026-05-23T02:34:56Z","created_by":"Yogthos","updated_at":"2026-05-23T03:09:57Z","started_at":"2026-05-23T02:45:40Z","closed_at":"2026-05-23T03:09:57Z","close_reason":"Closed","dependency_count":0,"dependent_count":0,"comment_count":0}

src/ui/mod.rs

@@ -689,6 +689,28 @@ pub async fn run_interactive(
     let mut was_reasoning = false;
     let mut todo_tools_enabled = false;
     let mut last_tool_name: Option<String> = None;
+    // The tool_call_id of the in-flight chamber (or the most-recent
+    // chamber that was closed without a matching ToolResult yet). Lets
+    // the ToolResult handler distinguish "this result belongs to the
+    // currently-painted chamber" (sequential / single-tool case) from
+    // "this result belongs to an earlier call whose chamber was
+    // displaced by a parallel sibling" (the dirge-jzj scenario).
+    //
+    // When parallel tool execution is enabled (the default per
+    // agent_loop/types.rs), the LLM emits N ToolCalls back-to-back and
+    // the agent_loop's `execute_tool_calls_parallel` fires
+    // ToolExecutionStart for ALL of them before any ToolExecutionEnd.
+    // Each new ToolCall passively closes the prior chamber. Completion
+    // order is whatever finishes first, so ToolResults arrive
+    // arbitrarily — most never match the currently-open chamber's id.
+    // Without this tracker, mismatched results either landed inside
+    // the wrong chamber (path a, body painted under another tool's
+    // banner) or as a `↳ first_line` trailer below an unrelated chamber
+    // (path b). The fix: when a result's id doesn't match the open
+    // chamber, paint a fresh complete chamber for THIS id below the
+    // current scroll position. Completion-order rendering, each tool
+    // gets its own correctly-labeled frame.
+    let mut last_tool_call_id: Option<String> = None;
     // Tracks whether a tool chamber TOP has been drawn but no matching
     // BOTTOM has been written yet. Used by the ask/alert handler to
     // close the in-flight chamber BEFORE rendering the ALERT box.
@@ -2136,6 +2158,7 @@ pub async fn run_interactive(
                         // `close_tool_chamber_if_open`.
                         close_tool_chamber_passive(&mut renderer, &mut last_tool_name, &mut tool_chamber_open)?;
                         last_tool_name = Some(name.to_string());
+                        last_tool_call_id = Some(id.to_string());
                         if agent_line_started {
                             renderer.write_line("", Color::White)?;
                             agent_line_started = false;
@@ -2201,6 +2224,61 @@ pub async fn run_interactive(
                         let max_chars = cfg.resolve_tool_result_max_chars();
                         let show_diff = cfg.resolve_show_edit_diff();
 
+                        // dirge-jzj: if the chamber on screen belongs to a
+                        // DIFFERENT tool call (parallel-execution race
+                        // where ToolResults arrive out of order, or a
+                        // newer ToolCall's TOP displaced this result's
+                        // chamber before the result arrived), paint a
+                        // fresh complete chamber for THIS id below the
+                        // current scroll position. Lets each result land
+                        // in its own correctly-labeled frame regardless
+                        // of completion order. The id-matches case (the
+                        // common sequential path) falls through to the
+                        // existing render paths below.
+                        if !id.is_empty()
+                            && last_tool_call_id.as_deref() != Some(id.as_str())
+                            && show_details
+                        {
+                            // Close whatever chamber is on screen first,
+                            // then paint a fresh TOP for this id. We
+                            // don't reuse the ToolCall handler's TOP-
+                            // paint code path because that fires from a
+                            // different event; the body of the new
+                            // chamber will land via path (a) below now
+                            // that tool_chamber_open=true.
+                            if tool_chamber_open {
+                                close_tool_chamber_passive(
+                                    &mut renderer,
+                                    &mut last_tool_name,
+                                    &mut tool_chamber_open,
+                                )?;
+                            }
+                            let (resolved_name, resolved_args) = tool_calls_buf
+                                .iter()
+                                .rev()
+                                .find(|e| e.id == id.as_str())
+                                .map(|e| (e.name.to_string(), e.args.clone()))
+                                .unwrap_or_else(|| (String::new(), serde_json::Value::Null));
+                            if !resolved_name.is_empty() {
+                                let upper = resolved_name.to_ascii_uppercase();
+                                let raw_value =
+                                    format_tool_banner_value(&resolved_name, &resolved_args);
+                                let raw_value = sanitize_output(&raw_value).into_string();
+                                let (frame_w, _) = chamber_widths(&renderer);
+                                let header =
+                                    fit_banner_header(&upper, &raw_value, frame_w);
+                                renderer.write_line("", Color::White)?;
+                                renderer.write_line(&header, c_tool())?;
+                                tool_chamber_open = true;
+                                last_tool_name = Some(resolved_name);
+                                last_tool_call_id = Some(id.to_string());
+                            }
+                            // If the call wasn't in tool_calls_buf (id
+                            // unknown — shouldn't happen post-ToolCall
+                            // but defensive), fall through to path (b)
+                            // trailer; we have no banner to paint.
+                        }
+
                         // on-tool-end is also fired by HookedToolDyn so the
                         // host doesn't re-dispatch it here.
 
@@ -2410,6 +2488,7 @@ pub async fn run_interactive(
                         // Clear after consuming so a future stray ToolResult
                         // can't be coloured with a stale tool name.
                         last_tool_name = None;
+                        last_tool_call_id = None;
                     }
                     AgentEvent::Done { response, tokens, cost } => {
                         was_reasoning = false;