- Displace overwritten destination into a temporary sibling directory during workspace renames.
- Restore the displaced file and clean up the temp directory if the main rename operation fails.
- applyWorkspaceEdit takes an onExecuted callback fired after each
filesystem mutation, so callers hold the executed prefix even when a
later op throws.
- applyWorkspaceEditWithLsp reconciles overlays/watchers for that prefix
best-effort before rethrowing the original apply error.
- applyWorkspaceEdit now returns { applied, executed }; ops skipped via
ignoreIfExists/ignoreIfNotExists are excluded from executed.
- applyWorkspaceEditWithLsp derives didClose/refresh/watched-file
notifications from executed ops, so a skipped rename no longer closes
the old URI overlay or emits phantom Deleted/Created events.
Skipped destination removal when an overwrite rename resolves both paths to the same inode (case-only rename on case-insensitive filesystems), so the source is no longer deleted before fs.rename runs.
Fixes#8373
Reject snippet-formatted or overlapping rename edits across all accepted URIs before applying any, so a later snippet edit cannot leave earlier files half-applied.
Fixes#8376
The willRenameFiles loop caught every non-abort, non-method-not-found
error into serverNotes and fell through to fs.rename, so a genuine
failure from a server that supports the request moved the path without
its semantic edits, leaving references dangling. Split client acquisition
from the request, track hard failures, and abort before any mutation when
a supporting server errors. Servers replying method-not-found are still
skipped without blocking.
Fixes#8380
The direct diagnostics loop caught non-abort server errors without recording them, so a file whose every applicable server failed produced an empty aggregate rendered as OK with success: true — a false-negative hiding a total diagnostics failure.
Track per-file and global success/failure counts: zero successful server responses now yields success: false with an explicit failure line, while partial success still surfaces diagnostics and names the servers that failed.
Fixes#8377
Propagated CreateFile, RenameFile, and DeleteFile options through workspace edit planning and enforced their overwrite, ignore, and recursive semantics during execution.
Fixes#8373
Advertised abort semantics because multi-file text edits are applied sequentially and cannot satisfy textOnlyTransactional.
Added initialization capability coverage and documented the fix.
Fixes#8375
- Added the `--external-thinking` CLI flag alongside model capability checks to gate external thinking tool availability.
- Updated Anthropic and Google transports to honor `forceReasoningOff` for native thinking-off controls.
- Renamed the `thoughts` property and parameter to `notes` across think fixtures, tools, and tests.
- Updated system prompt instructions and test suites to verify transport-specific thinking and tool activation.
- Added support for external thinking and forced reasoning disablement across AI provider options and request transformers.
- Implemented the private scratchpad think tool along with its renderer, system prompt rules, and schema configuration.
- Updated agent session management and SDK tools to support dynamic runtime activation of the think tool via the externalThinking setting.
- Added comprehensive unit tests covering reasoning fallbacks, tool activation, and rendering behavior.
- Define a centralized `USER_AGENT` constant in `@oh-my-pi/pi-utils` formatted as `omp/<version>`.
- Replace hardcoded and platform-specific user agent strings across AI providers, catalog scrapers, tools, and search providers with the unified `USER_AGENT`.
- Add unit tests for update-cli binary release distribution gating.
Key hints resolved modifier tokens through a static, platform-agnostic label map with no `super` entry, so on macOS the shipped `super+v` paste default rendered 'Super+V' (no such key on a Mac) and `alt` always rendered 'Alt' instead of 'Option'. The static /hotkeys navigation rows were hardcoded with macOS 'Option'/'Cmd' names on every platform, so Linux/Windows users saw 'Cmd+Left'.
Modifier labels are now platform-aware: on darwin `alt` renders 'Option' and `super` renders 'Cmd'; every other platform keeps 'Alt'/'Super'. The platform is resolved through a single seam (setKeyHintPlatform/keyHintPlatform) mirroring the TUI's setKittyProtocolActive, keeping hint output deterministic in tests without mutating process.platform. The static /hotkeys rows use the same convention and drop the macOS-only Cmd line-start/end fragments (which map to no binding) off darwin.
Fixes#8235
Address review feedback on #8186:
- add ty.toml to ty rootMarkers so ty.toml-only projects pass the
loadConfig marker gate (Codex P2, roboomp should-fix)
- add ty.toml to PYTHON_ROOT_MARKERS so project-local .venv/bin/ty
resolution runs for ty.toml-only projects (roboomp should-fix)
- add a ty.toml-only regression test covering both the detection
gate and the local venv bin resolution path
- move the changelog entry from the released [16.3.11] section into
[Unreleased] (roboomp should-fix)
Add ty (ty server) to defaults.json behind pyright/basedpyright/pylsp
in primary selection order and ahead of ruff (linter). ty uses the
generic LSP client path with no adapter code; settings.ty pass-through
works via the existing settings field.
- defaults.json: insert ty entry between pylsp and ruff
- lsp-regressions.test.ts: 4 tests (selection order for .py/.pyi,
auto-detect via $which + pyproject.toml, coexistence with ruff)
- CHANGELOG.md: Added entry under [Unreleased] referencing #4617Closes#4617
- Expanded web search and fetching providers with robust parsing, authentication storage integration, and response validation.
- Added support for new configurations including SearXNG safesearch, Cloudflare AI Gateway endpoints, and dynamic Firecrawl base URLs.
- Implemented comprehensive test suites covering error handling, content filtering, and provider-specific response behaviors.
Why:
Local timeout cleanup removes pending state but leaves the language server
working on an abandoned request.
Changes:
- Send $/cancelRequest after an issued request reaches its client timeout.
- Preserve the existing timeout error and pending-request cleanup.
Evidence:
- The workspace-readiness regression observes cancellation for the dropped
status request before polling succeeds.
Refs #8116
Why:
The module-level idle interval remains referenced after every client shuts
down and keeps short-lived hosts alive.
Changes:
- Stop the idle checker before draining active and pending clients.
- Preserve explicit reconfiguration through setIdleTimeout().
Evidence:
- A child-process regression probe exits after shutdown with the checker
configured.
Refs #8115
While the agent worked through a plan, every sub-todo rendered unchecked
no matter how far along the run was: the phase header highlighted, the
task rows below it looked untouched. Three separate causes, all on the
collapsed path that is the default view.
`selectCollapsedTodos` dropped every closed row while a phase held open
work, so finishing a task only ever *removed* a line — the panel never
rendered a checked box until the whole phase settled. That also made the
card's completion animation dead code: `details.completedTasks` drives a
14-frame strike reveal at 65ms with a component render per tick, against
a row the viewport had already discarded. The existing animation test
missed it by asserting on `expanded: true`.
The viewport now keeps the newest closed task as a checked lead row,
additive to the open-task cap so it never evicts open work, and the
strike sweep lands where users actually see it.
Second, the card gave a `done/total` count to every collapsed untouched
phase but not to the active one, so the phase being worked in was the
single phase reporting no progress. Extracted `formatPhaseProgress` and
put it on every phase header.
Third, the todo auto-clear (`tasks.todoClearDelay`, default 60s) armed
on any list holding a closed task and physically deleted those tasks
from the HUD's copy. An in-flight phase at `3/4` silently became `0/1`
sixty seconds later, fully-closed phases vanished, and stage roman
numerals renumbered off the filtered index — until the next `todo` call
restored the real snapshot. It now fires only once the whole list is
settled, which is the case the setting exists for; the walking viewport
already hides closed rows while work remains.
Progress counters also count closed tasks rather than only completed
ones. The viewport hides abandoned tasks too, so counting only
completions left a phase reading permanently stuck.
Preserved zero-width readiness and wait matches across the daemon wire protocol, and isolated malformed completion events from unrelated pending RPCs.
Fixes#7908
When an xd:// device is dispatched through the write tool, the outer
approval gate now consults tools.approval.<deviceName> before falling
back to tools.approval.write. This lets users scope allow/deny/prompt
to a single device mount without changing the blanket write tool policy.
The write tool's approval function returns { tier, policyKey: deviceName }
for xd:// device dispatches. resolveApproval uses the policyKey to look
up the user override on the device name, falling back to the invoking
tool's own policy when the device has none configured.
Adds:
- ToolApprovalDecision.policyKey field (optional, additive)
- policyKey-aware lookup in resolveApproval and requiresApproval
- Updated error messages naming the correct config key
- Unit tests for policyKey resolution and WriteTool integration
Fixescan1357/oh-my-pi#7923
The `cd <path> && ...` extractor matched everything up to the first `&&`
with a greedy regex, so a redirect or extra argument before the `&&` was
swallowed into the structured cwd. `cd /tmp 2>/dev/null && echo ok` became
cwd `/tmp 2>/dev/null`, which failed fs.stat and killed the command before
the shell ran.
Replace the regex with `extractLeadingCdTarget`, a quote/escape-aware
scanner in shell-tokenize.ts that captures exactly one path token and
bails (leaving the command for the shell) when anything else — a redirect,
extra argument, shell expansion, or a non-`&&` separator — precedes the
top-level `&&`.
Fixes#7883
can1357