- Added `renderSnapcompactFrames()` and `snapcompactFrameCount()` to @oh-my-pi/snapcompact for paging arbitrary text into PNG image blocks without dim-marker bookkeeping.
- Widened the agent loop's `transformProviderContext` hook to `(context, model) => Context` so per-request transforms can gate on the dispatch model's capabilities.
- Added `SnapcompactInlineTransformer` rendering the system prompt and large historical tool results as snapcompact frames on vision models: vision gate, per-provider image budgets, 3k-token floor, savings-margin gate, skip-last rule, and hash-keyed render caches swept to live tool calls.
- Added default-off `snapcompact.systemPrompt` and `snapcompact.toolResults` settings under a new Context → Experimental group, composed after secret obfuscation in `sdk.ts` so frames are built per-request and never persisted to session.jsonl.
- Added prompt stubs (`snapcompact-system-stub.md`, `snapcompact-system-frames-note.md`, `snapcompact-toolresult-note.md`) and unit tests covering frame paging, no-mutate guarantees, budget caps, gates, and render caching.
- Updated `settings-selector-memory-refresh.test.ts` to assert the cross-tab search banner's cursor rendering (`b▌`) instead of the old `Search: b` label when verifying Escape clears the search before closing the selector.
- Replaced the fixed `CALL_PREVIEW_MAX_LINES` cap in `render-utils.ts` with a dynamic `previewWindowRows()` (terminal rows minus a chrome reserve); `capPreviewLines` now returns lines untouched when expanded and windows them otherwise.
- Applied the cap in every render state for bash and ssh command previews — streaming and final render identically instead of snapping fully open on completion — and included the preview window in bash's render cache key so terminal resizes invalidate it.
- Capped eval cell code previews via the same window; previously cell code was never capped at all, with `ctrl+o` remaining the only way to uncap.
- Added stream-vs-final parity coverage in `bash-sixel-render.test.ts`, `ssh-render.test.ts`, and the new `eval-code-preview.test.ts`, and a changelog entry.
- Fixed `/dump` output (`session-dump-format.ts`) and the RPC `get_state` `dumpTools` payload (`rpc-mode.ts`) serializing Zod schema instances — leaking `def`/`shape` internals and stringified methods — by converting parameters through `zodToWireSchema` like providers receive.
- Fixed context-usage token estimation stringifying the Zod `def` tree, which overcounted tool schema tokens in the status line.
- Fixed tool-discovery indexing (`tool-index.ts`) returning empty `schemaKeys` for Zod tools, weakening BM25 ranking; keys are now recovered via wire conversion.
- Fixed the extension inspector panel rendering "(no arguments)" for Zod tools by reading properties off the converted wire schema.
- Added coverage in `tool-index.test.ts` plus new `context-usage.test.ts` and `session-dump-format.test.ts`, and a changelog entry.
- Added a `SettingsSearchHeader` banner to `SettingsSelectorComponent`, permanently mounted under the top border and rendering only while a cross-tab search is active.
- Typing now filters settings across every tab via `getSettingItemFilterText` + `fuzzyFilter`, with per-tab first-match ids so Tab jumps between tabs that have matches; non-matching tabs render muted and the pre-search tab is restored on exit.
- Moved the tab bar below the tab content, disabled `SettingsList`'s own type-to-search in favor of the selector-owned global search, and only advertised PgUp/PgDn in the hint when multiple sections exist.
- Wired the new `mutedTab` style into `getTabBarTheme` in `modes/shared.ts`.
Also normalizes setting labels/descriptions and carries the package changelog entries for this batch (adjacent unreleased bullets are not splittable per commit), restoring section placement disturbed by the two prior changelog hunks.
- Caught and logged npm `PluginManager.list()` rejections in `PluginSettingsComponent`, returning an empty plugin list so the settings tab no longer stays blank when plugin listing fails.
- Added tests covering Escape handling while lists are loading and marketplace rendering when npm listing rejects.
- Added `responsesLite`, `reasoningContext`, `clientMetadata`, and `onModerationMetadata` options.
- Changed Lite request transformation to default `reasoning.context` to `all_turns`.
- Changed request shaping to strip image `detail` fields and force `parallel_tool_calls` off in Lite mode.
- Fixed streaming behavior by excluding `client_metadata` from websocket append diffs and swallowing moderation metadata errors.
Unconfigured pi/designer now follows modelRoles.default before consulting the Gemini priority chain, matching the smol and slow fallback behavior while preserving priority defaults when no default is configured.
Added regression coverage and updated the coding-agent changelog.
Fixes#2336
When modelRoles.default points at another role alias (e.g. "pi/slow") and the inheriting role is unset, the resolver now recurses through resolveConfiguredRolePattern with a visited-role set so downstream one-layer expanders like completion-bridge's resolveTierModel see concrete model patterns instead of pi/<role>. Self-aliased defaults still collapse to the role's built-in priority chain.
Added regression coverage for both cross-role expansion paths.
Fixes#2336
When modelRoles.default points at pi/smol or pi/slow, unset matching roles now expand back to their built-in priority chain instead of returning the alias as a literal model pattern.
Added regression coverage for self-aliased smol and slow defaults.
Fixes#2336
Unconfigured pi/smol and pi/slow role aliases now inherit modelRoles.default before consulting the cloud-priority list, avoiding silent paid-provider routing for local-default setups.
Added resolver coverage for both roles and updated the coding-agent changelog.
Fixes#2336
Older GitHub CLI releases reject the optional issue stateReason JSON field. Retried single issue reads without it and stopped requesting it for issue listings.\n\nFixes #2333
- Added optional `requestModelId` to the `Model` interface for upstream wire overrides.
- Prioritized `model.requestModelId` when resolving request model IDs for Anthropic and OpenAI paths.
- Used `COPILOT_API_HEADERS` in Copilot discovery and policy calls with API version `2026-06-01`.
- Synthesized Copilot `-1m` long-context sibling models with dedicated upstream IDs and pricing data.
- Added `fanOutFileItems` plumbing through `resolveSearchPathItems`/`resolveExplicitSearchPaths` and `ToolScopeOptions` so plain-file entries can be scanned as explicit targets instead of being folded into an unrequested ancestor walk.
- Search tool now enabled `fanOutFileTargets` and deduplicated overlapping results by tracking a `path\0line` key before merging matches.
- Expanded multi-path tests for shared non-root ancestors, explicit `.git/config` targets, and overlap deduplication behavior.
Removed the dead settings selector submenu guard so Escape reaches the active settings or plugins child component. Added a regression test for closing an open settings submenu before the parent selector.
Fixes#2331
Programmatic `git push` operations should not follow annotated tags configured by a user's `push.followTags = true`. This setting can lead to push failures if the remote rejects tag creation (e.g., in PR-head forks), even if the branch update itself is valid.
Adding `--no-follow-tags` explicitly overrides this user setting, ensuring only the specified refspec is pushed and preventing rejections.
The format-on-write path sent a hardcoded {tabSize:3, insertSpaces:true}
on every textDocument/formatting request from two duplicated DEFAULT_FORMAT_OPTIONS
constants. Servers that honour tabSize for re-indent (yaml-language-server, the
common YAML/Kustomize/Flux case) reserialized 2-space files at a 3-space stride
on every write — exactly the corruption reported on Kubernetes/Flux YAML repos.
Replace both constants with a single resolveFormatOptions(filePath, content)
helper that layers, in order:
- .editorconfig (indent_style, indent_size, tab_width) via the new
getEditorConfigFormatting() helper in pi-utils — strict, no fallback.
- Indent sniffed from the in-memory content the agent is about to write
(first indented line decides spaces vs tabs; GCD of space-indent widths
fixes the unit).
- Hardcoded 2-space fallback. The previous 3-space stride was an unusual
default that actively damaged every file with a 2/4-space convention.
Tests cover the editorconfig and content-sniffing paths plus a direct
regression check that 2-space YAML stays 2-space (the issue's repro).
Fixes#2329
The root compose recorded seams last-writer-wins, so the status
container's new seam (mounted loader) overwrote the transcript's seam
whenever both reported in the same frame. With a volatile live block
(no commit-safe end) the boundary then moved below the transcript's
still-mutable rows and committed them as stale history — the same
duplication class this PR fixes for the loader, relocated upward.
Commits are prefix-only: the first child reporting a live region (plus
its own commit-safe extension) already bounds everything below it, so
the topmost seam now wins. Adds a two-reporter regression test and
documents the multi-reporter contract on NativeScrollbackLiveRegion.
The format-on-write path sent a hardcoded {tabSize:3, insertSpaces:true}
on every textDocument/formatting request from two duplicated DEFAULT_FORMAT_OPTIONS
constants. Servers that honour tabSize for re-indent (yaml-language-server, the
common YAML/Kustomize/Flux case) reserialized 2-space files at a 3-space stride
on every write — exactly the corruption reported on Kubernetes/Flux YAML repos.
Replace both constants with a single resolveFormatOptions(filePath, content)
helper that layers, in order:
- .editorconfig (indent_style, indent_size, tab_width) via the new
getEditorConfigFormatting() helper in pi-utils — strict, no fallback.
- Indent sniffed from the in-memory content the agent is about to write
(first indented line decides spaces vs tabs; GCD of space-indent widths
fixes the unit).
- Hardcoded 2-space fallback. The previous 3-space stride was an unusual
default that actively damaged every file with a 2/4-space convention.
Tests cover the editorconfig and content-sniffing paths plus a direct
regression check that 2-space YAML stays 2-space (the issue's repro).
Fixes#2329
The previous refactor dropped `if (this.aliasOf) return` into
consolidate(), which made `/memory enqueue` invoked from inside a
subagent a no-op. The old enqueue() body short-circuited only
forceRetainCurrentSession (via its own guard) but still flushed
extractions and slept on state.memory, which for aliased states points
at the parent's shared retain bank.
Moved the alias guard back out of consolidate(): forceRetainCurrentSession
keeps its own early-return (the subagent's transcript is the parent's
concern), so consolidate() runs the SQL-level flush+sleep on every
owned bank for both primary and aliased states. The lifecycle guard
stays in dispose() so disposing a subagent still does not flush, sleep,
or close the parent's memory.
Added a regression test: a /memory enqueue routed through the aliased
child state calls flushExtractions + sleepAllSessions(false) on the
parent's owned memory exactly once, and forceRetainCurrentSession runs
on the child (where its own guard returns early) but not on the parent.
Addresses #2327 review.
`/memory clear` (mnemopiBackend.clear) calls dispose() right before
removing the SQLite files. Running the consolidation pass first would
make a destructive clear spend tokens/time creating memories that get
deleted on the very next line, and a slow extraction could block the
clear path under llmMode: remote.
Added a `{ consolidate?: boolean }` option to
MnemopiSessionState.dispose; defaults to true so normal shutdown still
runs the SHMR/beam pipeline. mnemopiBackend.clear passes
`{ consolidate: false }` so the close path stays unsubscribe + close
only. Added a regression test that spies on forceRetainCurrentSession,
consolidate, flushExtractions, and sleepAllSessions to pin the new
behavior, while close still runs once per owned bank.
Addresses #2327 review.
MnemopiSessionState.dispose() now drains pending fact extractions and
runs sleepAllSessions on every owned bank before closing handles, and
AgentSession.dispose() awaits the result. This matches the manual
`/memory enqueue` slash command which was the only caller of the
consolidation pipeline.
Without the shutdown hook, episodic_memory, gists, consolidation_log,
graph_edges, and triples stayed empty for every deployment that never
typed `/memory enqueue|rebuild` — working memory accumulated forever
and long-term recall never formed.
Factored the consolidate step out of `mnemopiBackend.enqueue` into a
shared MnemopiSessionState#consolidate so the slash command and the
shutdown path share one implementation. Added two regression tests
verifying owned-bank consolidation order (flush -> sleep -> close, per
bank) and that aliased subagent dispose stays a no-op against the
parent.
Fixes#2320
- Extended `AgentTool.concurrency` to accept per-call resolver functions and resolved concurrency mode from each tool call, falling back to exclusive on resolver errors.
- Updated BashTool to schedule non-PTY calls as shared and PTY calls as exclusive so non-interactive bash calls can run in parallel within one message.
- Tracked in-use persistent shell sessions in the bash executor and routed overlapping calls on the same session key to isolated one-shot shells while preserving owner session availability.
A subagent's send await:true to Main during a blocking task spawn was a structural deadlock: deliverIrcMessage queues mid-turn messages as step-boundary asides, but Main's next boundary requires the sender's own batch to finish, so the sender always burned the full irc.timeoutMs. Awaited sends now pass expectsReply through IrcBus.send; a mid-turn recipient with async.enabled off generates an ephemeral no-tools reply via runEphemeralTurn, records an irc:autoreply aside in its own history, and delivers it back over the bus with replyTo threading so the sender's waiter resolves.
PR #2297's new test interleaved fixture writes with loader calls in one temp
dir; Bun's resolver caches directory entries process-wide, so tool.ts written
after loadExtensions() resolved extension.ts was invisible to the subsequent
dynamic import on macOS. Write all fixtures up front.
- Removed the running count from the status meta list so only completed, failed, and cancelled counts are shown.
- Updated the waiting message to use singular/plural job labels and simplify wording when all jobs are running.
can1357