ur-agent 1.44.1 → 1.44.3

package/CHANGELOG.md

@@ -1,5 +1,62 @@
 # Changelog
 
+## 1.44.3
+
+- Make thinking visually distinct from answers: thinking blocks are labeled
+  "model reasoning to itself — not the answer" (dim italic, left-bordered when
+  expanded); answer text carries an accent-colored ⏺ marker.
+- Pin the live task panel: it stays visible above the prompt while the agent
+  works (previously hidden during turns), with real-time status icons
+  (✔ done, ■ in progress, □ pending, ✘ failed, ⚠ skipped). `showExpandedTodos`
+  now defaults on; ctrl+T still toggles.
+- Add tasks on request: saying "add to your tasks …" now creates the task
+  immediately, even mid-turn, and it appears live in the pinned panel.
+- Add `/undo`: restores the most recently edited file to its content from
+  before the last turn's edits (deletes a file the last edit created). Full
+  checkpoint restore remains `/rewind`.
+- Reject shell commands with unterminated quotes before execution, with a
+  diagnostic that names the defect, suggests the quoted-heredoc fix, and
+  tells the model not to retry the identical command.
+- Add Concise, JSON-strict, Debug-verbose, and Release-notes output styles.
+- Ollama: drop text-form tool calls that duplicate native ones (no more
+  double-executed Writes); tolerate and strip up to two hallucinated extra
+  keys on bare-JSON Write/Edit calls instead of leaking them as prose.
+- Fix the test suite: repaired two committed test files that failed or
+  poisoned other suites via global module mocks — the suite is now fully
+  green (1134 pass, 0 fail).
+- TypeScript now covers test/ (fixed all 171 surfaced errors) and 544
+  compiled files no longer end in megabyte-scale inline sourcemap comments —
+  VS Code diagnostics and long-line problems are gone.
+- Repo hygiene: untracked committed local state (.ur state files, FUSE
+  artifact, stray yarn.lock) and added ignore rules; added UR.md project
+  memory so the agent learns the repo across sessions.
+- `Task` status schema now includes `failed` and `skipped`, matching what
+  crew/workflow runs produce and the task panel renders.
+
+## 1.44.2
+
+- Fix Ollama streamed tool-call accumulation: Ollama streams each completed
+  tool call in its own chunk, but the merge logic overwrote call N-1 with
+  call N, collapsing multi-call turns (e.g. several `Write` calls scaffolding
+  a test suite) into just the last call. Calls now append; string argument
+  fragments concatenate; empty argument resends no longer clobber good
+  arguments; cumulative resends stay idempotent.
+- Repair almost-JSON tool-call arguments instead of silently emptying them.
+  Local models routinely emit raw newlines inside JSON string values,
+  markdown fences, and trailing commas; strict parsing collapsed the whole
+  input to `{}`, which surfaced as `InputValidationError: required parameter
+  \`file_path\` is missing` and trapped the model in a retry loop. A lenient
+  parser (`parseToolInputJsonLenient`) now repairs these across the Kimi
+  marker parser, the bare-JSON text parser, the Ollama input parser, and
+  streamed tool-input normalization. Repairs are tracked via the
+  `tengu_tool_input_json_repaired` event; only genuinely hopeless input still
+  falls back to `{}`.
+- Warn (once per model, in debug logs) when an Ollama model does not
+  advertise the `tools` capability, since tool definitions are then silently
+  dropped. In that mode the system prompt now includes a concise instruction
+  telling the model to emit single-line bare-JSON tool arguments — the exact
+  format the text parser recovers — instead of leaving it to guess.
+
 ## 1.44.1
 
 - Fix task board rendering: finished, failed, and skipped tasks now render as