ToolPiper MCP Server

All 468 tools.

Every built-in tool in the ToolPiper MCP server, across 14 categories. Most run free on-device; the rest carry a badge for the tier that unlocks them.

Enforced at the API: a tool below your tier returns a 402 upgrade prompt. Tap a tier to filter.

See what each tier unlocks →

468 of 468 tools

Core

LLM inference, voice chat, TTS, STT, models, embeddings, OCR

audio_speakProConvert text to speech and save as a WAV file.

audio_transcribeProTranscribe an audio file (WAV, MP3, M4A, FLAC) to text using on-device speech recognition.

audio_voice_cloneProGenerate speech mimicking a voice from a reference audio sample.

chatRun a prompt through a local LLM on this machine via ToolPiper.

claude_code_installInstall the bundled claude-tp helper so the user can run `claude-tp` in their terminal to launch Claude Code with ToolPiper as the inference backend and ToolPiper's MCP tools mounted.

claude_code_statusRead-only diagnostic for the claude-tp helper installed by claude_code_install.

claude_code_uninstallReverse claude_code_install: removes the ~/.local/bin/claude-tp symlink, deletes the helper's config file, and revokes the inference bearer.

endpoint_listList ToolPiper inference endpoints (LLM providers + models).

endpoint_recommendSuggest the best inference backend for the current conversation.

endpoint_setSwitch the inference backend ToolPiper uses.

endpoint_statusShow the currently active inference backend for the caller.

model_checkpoint_addRegister the user's own pinned HuggingFace weights against a coded native media architecture family and start the download.

model_downloadStart downloading (or registering) a model, LoRA, VAE, or embedding from any source.

model_download_cancelCancel an in-flight model download by its download_id (from model_download_list or the response of model_download).

model_download_listList active model downloads with their progress and download_id values.

model_listList AI models available in ToolPiper.

model_loadLoad a model that is already downloaded into memory for inference.

model_repo_infoFetch metadata, the full file list, a recommended variant, and (for community sources) the license and trigger words for a specific candidate.

model_searchFind a model, LoRA, VAE, or embedding by free-text query across a source.

model_unloadUnload a model from memory to free RAM/VRAM.

provider_listList the provider settings: which provider each feature uses (the language model, the image, video, 3D and splat generators, voice chat and its fallback, segmentation, vision grounding), where that provider runs (this Mac, a server on the network, or a cloud provider whose messages leave the Mac), what the feature uses while a setting is unset, and the providers that may fill each setting, with any that cannot marked and the reason.

provider_setChange which provider a feature uses: set one provider setting from provider_list to one of that setting's choices, or omit provider to clear it so the feature uses its default.

system_resourcesGet a live snapshot of Mac hardware resources — total/used/available RAM, GPU memory (Metal allocated + max, device utilization), GPU thermal state (nominal, fair, serious, critical), ANE activity (which backends are using the Neural Engine), and chip info (name, memory bandwidth).

system_statusCheck if ToolPiper is running and get system status — loaded models, active backends, GPU thermal state, available RAM, and browser CDP connection state (connected, pageUrl).

text_embedGenerate vector embeddings for text, entirely on-device.

text_translateTranslate text between languages using Apple's on-device Translation framework.

tool_docsGet the full description and typed input schema for named tools.

tool_searchFind the right tools for a task by describing what you want to do in natural language.

vision_camera_captureProCapture a single still frame from the Mac's webcam and return it as a base64-encoded PNG (≤1280 px on the longest edge).

vision_ocrProExtract text from an image using Apple Vision OCR.

voice_chat_memory_addStore a persistent fact about the user that voice chat can reference in future sessions.

voice_chat_memory_clearWipe all persistent facts stored via voice_chat_memory_add.

voice_chat_session_resetClear the active voice chat conversation history.

voice_chat_settings_getRead the current voice chat configuration — endpointConfigId, TTS voice, TTS speed, tools-enabled flag, and session timeout.

voice_chat_settings_setUpdate voice chat settings; omitted fields keep their current values.

Analysis

Image/text analysis, RAG, PDF extraction, upscaling

asset_deleteProPermanently delete an Asset record by id and reclaim the artifact bytes it solely owns.

asset_describeProChat-to-feature reference transform.

image_adjustStudioStudio image edit.

image_analyzeProAnalyze an image using Apple Vision.

image_black_and_whiteStudioStudio image edit.

image_blend_maskedStudioStudio transform — apply ANY image edit to just one REGION of an image.

image_conditioningProStudio transform — structural-conditioning producer.

image_cropStudioStudio image edit.

image_curvesStudioStudio image edit.

image_cutoutStudioStudio transform.

image_editStudioStudio transform (image-to-image), REGION-SCOPABLE.

image_erase_regionStudioStudio transform — GENERATIVE ERASE.

image_fillProStudio image edit.

image_filterStudioStudio image edit.

image_flipStudioStudio image edit.

image_free_transformProStudio image edit.

image_generateStudioStudio transform.

image_gradientProStudio image edit.

image_gradient_mapProStudio image edit.

image_grayscaleStudioStudio image edit.

image_importProStudio ingress (Apple typed-media rich import).

image_inpaintStudioStudio transform — GENERATIVE FILL of one region.

image_invertStudioStudio image edit.

image_levelsStudioStudio image edit.

image_mask_from_shapeStudioSelection without a model: a rectangle, ellipse or polygon (a lasso is a polygon) over an `image` becomes a NEW `mask` at its native size, with optional 'feather' and a set operation against an existing mask of the same image.

image_mask_refineStudioStudio selection edit.

image_outpaintProStudio transform — CANVAS EXTEND.

image_perspectiveProStudio image edit.

image_posterizeProStudio image edit.

image_remove_subjectStudioStudio transform.

image_resizeStudioStudio image edit.

image_rotateStudioStudio image edit.

image_segmentStudioStudio transform — the region authority.

image_straightenStudioStudio image edit.

image_strokeStudioONE brush stroke on ONE target, minting a NEW Asset of the target's type.

image_textProStudio image edit.

image_thresholdProStudio image edit.

image_to_splatStudioStudio transform.

image_to_worldStudioStudio transform.

image_transformStudioTransform or edit an image file on disk — resize (including shrinking and making images smaller), crop, rotate an image, or convert image format between PNG, JPEG, HEIC, WebP, TIFF.

image_upscaleStudioUpscale a still image using CoreML super-resolution.

image_upscale_benchmarkStudioMeasure PiperSR upscale performance on this machine.

image_upscale_neuralStudioStudio transform — GENERATIVE detail upscale.

pdf_extractExtract text, per-page text, images, or metadata from a PDF file.

qr_generateProGenerate a scannable QR code IMAGE.

rag_collection_listProList all RAG (Retrieval-Augmented Generation) document collections in ToolPiper.

rag_ingestProTrigger re-indexing of a RAG collection.

rag_queryProSearch local knowledge base collections using semantic similarity (vector + BM25 hybrid search).

support_searchAnswer a question ABOUT a Piper app itself from the curated help corpus that ships inside ToolPiper — ToolPiper, VisionPiper, AudioPiper, MediaPiper and MobilePiper, plus plans, pricing and what an upgrade adds.

text_analyzeAnalyze text using Apple's NLP framework.

video_importStudioStudio ingress (byte-to-typed-Asset video import).

Browser

AX-native browser automation, scraping, assertions

browser_actionProDispatch a browser interaction.

browser_assertProAssert a condition on the current browser page.

browser_autofillProTrigger Chrome's autofill on form fields.

browser_connectOwn the Chrome CDP connection lifecycle.

browser_consoleProGet browser console messages (log, warn, error) and optionally network request logs.

browser_coverageProMeasure JavaScript and CSS code coverage.

browser_detectProDetect the frontend framework, meta-framework, CMS, state libraries, router, and dev/prod mode on the current page.

browser_dialog_handleProResolve the pending JavaScript dialog.

browser_element_getProRead one element on the current page: its tag, id, classes and attributes, accessible role and name, border box, whether it is visible and in the viewport, a short text excerpt, its child count, a summary of its parent, and the framework component that owns it on a development build (its name, and its source file and line where the framework records them).

browser_emulateProEmulate device and environment conditions on the active Chrome tab.

browser_evalProExecute JavaScript in the browser page and return the result.

browser_interceptProIntercept and mock network requests.

browser_locatorProGiven an element on the current page, return the canonical, ranked set of stable accessibility selectors to use in saved tests.

browser_networkProInspect network traffic in the browser with full request/response detail.

browser_page_closeProClose a Chrome tab by its id (from browser_page_list).

browser_page_listList the open Chrome tabs the CDP debugger sees, with each tab's id, URL, title and lease: yours when this session's workspace holds it, held when another workspace does.

browser_page_newProOpen a new Chrome tab, optionally navigating to a URL.

browser_page_selectProSwitch the active Chrome tab the other browser_* tools operate on.

browser_pdfProRender the current browser page to a PDF file and return its path as a resource link.

browser_performanceProMeasure and audit browser performance.

browser_recordProRecord user interactions as action events (clicks, fills, navigations) for test creation — not screen video (use video_record for that).

browser_snapshotGet a structural snapshot of the current browser page via Chrome's accessibility engine.

browser_source_getProRead one stylesheet's or script's text from the current page, by the id browser_source_list returns or by its URL, as a window of lines: startLine (default 1) and lineCount (default 200, at most 2,000).

browser_source_listProList the code the current page has loaded: every stylesheet and script, external and inline, each with an id for browser_source_get, its kind, its URL (an inline one names the element holding it instead), its length in characters, and its source map URL when it declares one.

browser_storageProRead and write browser storage: cookies, localStorage, sessionStorage.

browser_style_getProExplain one element's styling on the current page.

browser_viewport_resizeProSet the viewport width and height for the active Chrome tab, and optionally the device pixel ratio it renders at.

browser_webauthnProTest WebAuthn/passkey flows with a virtual authenticator.

browser_webmcpUse the tools a website itself provides through WebMCP, where a page registers agent-callable tools on document.modelContext.

http_requestProMake a raw HTTP request to any URL — returns the HTTP status code, response headers, and raw response body (JSON, text, or bytes).

http_request_replayProReplay a discovered API endpoint with optional parameter overrides.

web_api_discoverProPassively analyze already-captured browser network traffic to discover API endpoints.

web_scrapeProFetch a URL in a real Chrome browser and return rendered content.

youtube_transcriptProFetch the transcript (captions) of a YouTube video.

Filesystem

File operations, archives, Git, workspace search

archive_createProBundle files and folders into a zip / tar / tar.gz archive.

archive_extractProUnpack a zip / tar / tar.gz archive back into files.

archive_listProList entries inside a zip / tar / tar.gz archive without extracting.

code_searchProgrep INSIDE source files.

doc_exportExport a doc, or a folder of docs as one book, to a file of another format.

doc_importBring a file of another format into Writing as a doc the person can edit.

file_copyProCopy a file or a directory tree to a new path, creating missing parent directories.

file_createProCreate a new file with optional initial content.

file_create_directoryProCreate a directory, creating missing parent directories as needed.

file_deleteProPermanently delete a single file from the filesystem.

file_delete_directoryProDelete a directory.

file_editProReplace an exact, unique span of text in an existing file — the surgical alternative to rewriting the whole file with file_write.

file_list_directoryProList a directory under `entries` (name, path, type, size, mtime epoch-seconds, is_symlink, kind UTI).

file_moveProMove or rename a file or directory.

file_pick_directoryProOpen a native macOS file/folder picker dialog and return the selected path with isDirectory flag.

file_readProRead the contents of a text file as a UTF-8 string.

file_writeProWrite content to a file, creating parent directories if needed.

git_checkoutProSwitch to a branch, tag, or commit hash.

git_commitProCreate a git commit with the currently staged changes.

git_diffProShow file diffs in a git repository.

git_logProShow recent commit history for a git repository.

git_pushProPush commits to a remote repository.

git_statusProShow the working tree status of a git repository.

system_run_commandProExecute a shell command and stream stdout/stderr in real time.

workspace_addProGrant the agent access to a directory this session by pinning it to the workspace access list.

workspace_searchProFind files and folders by filename or path prefix.

Capture

Screenshots, camera capture, audio recording

audio_recordProRecord audio — capture N seconds of audio to a WAV file.

vision_color_pickProEyedropper color picker — pick a color from anywhere on screen and get its hex code.

vision_recordStudioRecord the screen to a video clip — a screen recording, screencast, or capture of what is happening on a display or in a region of it, through VisionPiper — for a fixed number of seconds, or until stopped (duration_seconds=0, then command=stop).

vision_region_getProRead the on-screen capture region VisionPiper has set: its rectangle in points relative to its display (top-left origin), the display_id, the generation counter that moves on every change, whether the border indicator is showing, and whether a screen recording is running right now.

vision_region_setProSet the on-screen capture region: the one rectangle VisionPiper draws a border around and that vision_record records and vision_screenshot mode=region can capture.

vision_screenshotProTake a screenshot — capture the screen as a PNG image.

Testing

PiperTest CRUD and execution

test_deleteMaxDelete a saved PiperTest by ID.

test_exportMaxExport a PiperTest to Playwright or Cypress code.

test_getMaxGet a full PiperTest by ID with all steps, AX paths, element metadata, and execution context.

test_listMaxList saved PiperTests with metadata (id, name, step count, last run status).

test_runMaxExecute a PiperTest via CDP at native speed (~10-50ms/step).

test_saveMaxCreate or update a grouped PiperTest.

Motion

Pose estimation, real-time skeleton streaming

pose_detectStudioDetect human body poses in an image.

pose_format_listStudioList available pose detection backends, renderers, and output formats.

pose_stream_startStudioStart the real-time pose streaming WebSocket server (on-demand) and return connection details.

vision_stream_startProProcess a live WebSocket video feed with vision processors (pose, OCR, classify, upscale, faces, barcodes).

vision_stream_stopProStop a running stream processing session.

Outreach

GitHub, Search Console, HN, Reddit, content queue

github_activityProFetch recent activity for a GitHub repo: commits, releases, merged PRs, star/fork counts.

github_compareProCompare two Git refs (branches, tags, or SHAs) in a GitHub repo.

github_repo_listProList GitHub repositories accessible by the configured PAT (up to 100, sorted by last push).

gsc_analyticsQuery Google Search Console search analytics: clicks, impressions, CTR, position.

gsc_inspectInspect URLs in Google Search Console: indexed?, canonical URL, last crawl date, mobile usability, rich results.

gsc_site_listList all Google Search Console properties the account has access to.

gsc_sitemap_listList sitemaps for a GSC property with error/warning counts, last downloaded date, and indexed URL count.

hn_searchProSearch Hacker News stories and comments via Algolia.

hn_trending_listProGet trending stories from Hacker News right now.

queue_addProAdd content as a draft to the outreach queue.

queue_listProList drafts in the outreach content queue.

queue_publishProMark a queued draft as published.

reddit_postProSubmit a new post or reply on Reddit.

reddit_searchProSearch Reddit posts by keyword, optionally scoped to a subreddit.

web_searchGeneral-purpose open-web search.

Search

SERP search, autocomplete, SEO rank tracking

serp_autocompleteProKeyword discovery via Google Suggest — returns the real autocomplete queries people search for a seed term.

serp_rank_checkProFind where specific domains rank in the live Google or Bing results for a query.

serp_searchProCapture the live Google or Bing search-results page for a query via a real browser.

Sieve

MediaPiper sieve inspection, testing, healing

sieve_healMaxDiscover or repair full-size image URL patterns for a domain's thumbnails using a local LLM.

sieve_inspectMaxGet full sieve detail for a specific domain: rule definition, canary URL pairs, mutation ops, confidence history, and health state.

sieve_listMaxList all cached image discovery sieves from the MediaPiper browser extension.

sieve_setMaxEdit MediaPiper site rules.

sieve_testMaxDry-run: apply the cached sieve for a domain to a specific URL.

System

204

macOS system actions — audio, display, windows, Finder, clipboard, and more

automation_createMaxArm an automation in a project (the default project when none is named) that fires on a clock: a one-shot instant, a repeating interval, or a cron expression.

automation_deleteMaxDelete an automation for good, with its run history.

automation_getMaxRead one automation, including its recent firings: the outcome, the detail, the chat a chat firing created, and how many missed firings each one coalesced.

automation_listMaxList automations with their project, trigger, action, enabled state, next firing and last outcome, optionally only one project's, plus any automation file that could not be read.

automation_run_nowMaxFire an automation now, as the principal that armed it, without waiting for its trigger.

automation_updateMaxChange an automation's name, enabled state, trigger or action.

autopilot_history_queryProRetrieve prior-iteration history (summaries, decisions, dead-ends, full diffs) for YOUR Autopilot loop's cold store.

autopilot_resumeProAnswer a parked escalation and resume a paused Autopilot loop.

autopilot_startProKick off a goal-conditioned autonomous agent loop that drives itself across fresh headless contexts until the goal is met or a real escalation occurs.

autopilot_statusProReport the current progress, run status, criteria gap, and any parked escalation for an Autopilot loop.

autopilot_stopProCancel a running (or parked) Autopilot loop and mark it terminal (STOPPED).

conversation_createCreate an empty chat in a project (the default project when none is named).

conversation_deleteDelete a chat for good.

conversation_getRead one chat: its title, project, open workbench tabs, active tab, whether the pane is open, its latest run and revision, and its last message.

conversation_listList chats with their id, title, project, open workbench tabs, latest run state and revision, most recent first, optionally only one project's, plus any chat file that could not be read.

conversation_moveMove a chat to another project.

conversation_pane_setSet what a chat's pane shows.

conversation_pinPin a chat to the top of its project's sidebar.

conversation_renameGive a chat a new title.

conversation_unpinUnpin a chat, so it sorts with the project's other chats again.

conversation_workbench_closeClose a workbench's tab in a chat's pane.

conversation_workbench_openOpen a workbench of the chat's project as a tab in the chat's pane, after the last tab.

job_cancelCancels a still-running (or pending) async job by id: flips it to `cancelled` and stops the underlying work.

job_resultReturns the result of a COMPLETED async job: the `outputAssetId` of the Asset it produced (plus the resolved Asset record with its provenance).

job_statusReturns the current state of an async job by id: `pending`, `running` (with a progress percent), `completed`, `failed`, or `cancelled`.

project_createCreate a project: a named container for chats and workbenches with shared instructions, a default model endpoint, the folders its chats read without asking, and materials (files and folders) its chats draw knowledge from.

project_deleteDelete an empty project.

project_getRead one project: its name, instructions, default endpoint, folders, materials, the folders and materials still pending the person's approval, and its revision.

project_listList every project with its id, name, revision and last update, including the default project, plus any project file that could not be read.

project_updateChange a project's name, instructions, default endpoint, folders or materials.

system_accessibility_contrast_setEnable or disable Increase Contrast.

system_accessibility_reduce_motion_setEnable or disable Reduce Motion.

system_accessibility_stateRead current macOS accessibility state: VoiceOver, zoom, reduce motion, increase contrast.

system_accessibility_voiceover_setEnable or disable VoiceOver.

system_accessibility_zoom_setEnable or disable screen Zoom.

system_app_assert_attributeAssert a named AX attribute (e.g. AXValue, AXTitle) on a native macOS app element equals expected.

system_app_assert_countAssert the count of matching native macOS app elements equals expected.

system_app_assert_enabledAssert a native macOS app element is enabled (not greyed-out).

system_app_assert_focusedAssert a native macOS app element has keyboard focus.

system_app_assert_hiddenAssert a native macOS app element is hidden or absent.

system_app_assert_textAssert a native macOS app element's text equals the expected value.

system_app_assert_visibleAssert a native macOS app element is visible.

system_app_circuit_resetReset the per-session on-device grounding circuit breaker.

system_app_clickProClick a native macOS app element.

system_app_confirmProPress the default/confirm button in a dialog or sheet of a native macOS app.

system_app_fillProFill a text field in a native macOS app.

system_app_listList all running macOS applications with bundle IDs, names, PIDs, and active/hidden state.

system_app_pressProPress a control in a native macOS app (button, checkbox, segmented control).

system_app_resolveLocate a native macOS app element by natural-language description when no selector will address it — an icon-only button, a custom-drawn control, a canvas label.

system_app_scrollProScroll a native macOS app to bring an off-screen element into view.

system_app_set_valueProSet the value of a slider, stepper, or other AXValue-bearing element in a native macOS app.

system_app_show_menuProOpen the context menu or popup menu on a native macOS app element.

system_app_snapshotGet the accessibility tree of a running native macOS app — the native analogue of browser_snapshot.

system_appearance_accent_setSet the system accent color.

system_appearance_highlight_setSet the system highlight color (used for selected text).

system_appearance_reduce_transparency_setEnable or disable Reduce Transparency.

system_appearance_sidebar_size_setSet Finder/Mail sidebar icon size.

system_appearance_stateRead current macOS appearance state: accent color, highlight color, transparency, sidebar size.

system_audio_device_listList all available audio output devices on this Mac.

system_audio_output_setSwitch the active audio output device by name or ID.

system_audio_stateRead current audio state: volume level, mute status, current output device.

system_batchProExecute multiple actions in a single call.

system_bluetooth_connectConnect to a paired Bluetooth device by MAC address.

system_bluetooth_device_listList paired Bluetooth devices with names, MAC addresses, and connection state.

system_bluetooth_disconnectDisconnect a connected Bluetooth device by MAC address.

system_bluetooth_setTurn Bluetooth on or off.

system_bluetooth_stateRead Bluetooth state: power on/off, paired device count.

system_calendar_event_createCreate a calendar event.

system_calendar_event_deleteDelete a calendar event by ID.

system_calendar_event_listList calendar events in a date range.

system_calendar_event_updateUpdate a calendar event by ID.

system_calendar_listList all macOS calendars.

system_clipboard_clearProClear the persistent clipboard history (pinned items survive).

system_clipboard_deleteProDelete a clipboard item by ID.

system_clipboard_getProGet a single clipboard item by ID.

system_clipboard_historyProBrowse the persistent clipboard history.

system_clipboard_pinProPin a clipboard item so it survives clears.

system_clipboard_readProRead the current clipboard contents (most recent item).

system_clipboard_restoreProRestore a clipboard item to the active clipboard.

system_clipboard_searchProFuzzy-search the persistent clipboard history (text + OCR).

system_clipboard_unpinProUnpin a pinned clipboard item.

system_clipboard_writeProWrite text, an image, or file paths to the clipboard.

system_contact_createCreate a new contact.

system_contact_getGet a contact by ID.

system_contact_searchSearch macOS contacts by name.

system_contact_updateUpdate an existing contact by ID.

system_default_handler_getGet the default app handler for a UTI type or URL scheme.

system_default_handler_setSet the default app handler for a UTI type or URL scheme.

system_defaults_readRead a macOS UserDefaults key.

system_defaults_writeProWrite a macOS UserDefaults key.

system_desktop_hot_cornersRead or set hot corner actions.

system_desktop_screensaver_startStart the macOS screen saver immediately.

system_desktop_stage_manager_setEnable or disable Stage Manager.

system_desktop_wallpaper_getRead the current desktop wallpaper path for one display.

system_desktop_wallpaper_setSet the desktop wallpaper from an image file, on one display.

system_display_arrangeSet a display's origin in the desktop arrangement, in global points: the main display is at (0, 0), y down, and system_display_list reports every display's current origin in the same space.

system_display_brightness_setSet brightness on any display that takes it: the built-in panel, and external monitors that speak DDC/CI.

system_display_dark_mode_setTurn dark mode on or off.

system_display_listList every online display, left to right: displayId (the id every system_display_* verb takes), name as System Settings shows it, size in points and pixels with scale, refresh rate, origin in global points, physical size, brightness where adjustable, and main / built-in / mirroring / placement flags (left of main, right of main, ...).

system_display_mirror_setEnable or disable display mirroring.

system_display_mode_listList a display's supported modes grouped by resolution with their refresh rates, current mode marked.

system_display_night_shift_setTurn Night Shift on or off.

system_display_resolution_setSet the display resolution.

system_display_rotateRotate a display to 0, 90, 180, or 270 degrees.

system_display_stateRead display state: dark mode, Night Shift, True Tone, mirroring, brightness where a display exposes it, and a `displays` array with every online display's id, name, size in points and pixels, position, refresh rate and placement relative to the main display.

system_display_true_tone_setTurn True Tone on or off.

system_dock_app_addAdd an app to the Dock by bundle ID.

system_dock_app_removeRemove an app from the Dock by bundle ID.

system_dock_autohide_setEnable or disable Dock auto-hide.

system_dock_magnification_setEnable or disable Dock magnification and set its size.

system_dock_position_setSet Dock screen position.

system_dock_size_setSet Dock tile size in pixels.

system_dock_stateRead current Dock state: position, tile size, auto-hide, magnification.

system_finder_infoProGet structured file metadata (path, name, type, size, mtime, kind/UTI, permissions, owner, tags) as JSON for a file or folder.

system_finder_openOpen a file or folder, optionally with a specific app.

system_finder_revealReveal a file or folder in Finder (selects the item in its containing folder).

system_finder_tagsRead or set Finder tags on a file.

system_finder_trashMove a file or folder to the Trash (recoverable).

system_finder_trash_emptyPermanently delete all items currently in the Trash.

system_focus_profile_listList configured macOS Focus profiles.

system_focus_setEnable or disable Focus / Do Not Disturb.

system_focus_stateRead current Focus / Do Not Disturb state.

system_infoGet system info: hardware, OS version, CPU, memory.

system_input_clickProSimulate a mouse click at screen coordinates.

system_input_dragProSimulate a mouse drag from (fromX, fromY) to (toX, toY).

system_input_keyProSimulate a key press with optional modifiers.

system_input_moveProMove the mouse cursor to screen coordinates.

system_input_scrollProPost a raw wheel event at a screen point.

system_input_typeProSimulate typing text character-by-character.

system_location_currentGet current GPS location.

system_location_statusCheck location services status.

system_lockLock the screen.

system_logoutLog out the current user.

system_media_nextSkip to the next track on the currently playing media app.

system_media_now_playingRead current media now-playing info (title, artist, album, position) across all macOS playback apps (Music, Spotify, YouTube, etc.).

system_media_play_pauseToggle play/pause on the currently playing media app.

system_media_prevSkip to the previous track on the currently playing media app.

system_mute_setMute or unmute audio output.

system_network_dnsRead the currently configured DNS servers.

system_network_interface_listList all network interfaces with names, types, IP addresses, and status.

system_network_joinJoin a Wi-Fi network by SSID.

system_network_pingCheck connectivity to a hostname or IP address.

system_network_scanScan for visible Wi-Fi networks and return their SSIDs and signal strength.

system_network_ssidRead the SSID of the currently joined Wi-Fi network.

system_network_stateRead overall network state: Wi-Fi on/off, current SSID, IP, connectivity.

system_network_wifi_setTurn Wi-Fi on or off.

system_notification_clearClear all delivered notifications from Notification Center.

system_notification_postPost a local macOS notification immediately.

system_notification_scheduleSchedule a local macOS notification to fire after a delay.

system_observeGet a composite snapshot of the entire Mac state in one call.

system_power_prevent_display_sleepPrevent the display from sleeping while the Mac stays awake.

system_power_prevent_sleepPrevent the Mac from sleeping (caffeinate).

system_power_stateRead battery level, charging status, and thermal state.

system_process_frontmostGet the currently-frontmost app.

system_process_infoGet info about a specific process by PID or bundle id.

system_process_killProTerminate a process by PID.

system_process_launchProLaunch a macOS app by bundle id or display name.

system_process_listList running processes / apps.

system_reminder_completeMark a reminder complete or uncomplete.

system_reminder_createCreate a reminder.

system_reminder_deleteDelete a reminder by ID.

system_reminder_listList reminders, optionally filtered by list.

system_reminder_list_listList all macOS reminder lists.

system_restartRestart the Mac.

system_settings_openOpen System Settings, optionally to a specific pane.

system_shortcut_listList all installed macOS Shortcuts.

system_shortcut_runRun a macOS Shortcut by name.

system_shutdownShut down the Mac.

system_sleepPut the Mac to sleep.

system_spaces_deleteProDelete a Mission Control space by ID.

system_spaces_listProList all Mission Control spaces/desktops with IDs and indices.

system_spaces_switchProSwitch to a Mission Control space by 1-based index.

system_spaces_window_moveProMove a window to a different Mission Control space.

system_storage_ejectEject a removable volume by mount point.

system_storage_usageCheck disk usage for a filesystem path.

system_storage_volume_listList mounted volumes (internal disks, external drives, network shares).

system_timer_cancelCancel a specific running timer by ID.

system_timer_clearCancel all running timers.

system_timer_listList all currently-running timers with remaining time.

system_timer_startStart a countdown timer that fires a macOS notification when it elapses.

system_undoProUndo the most recent reversible action: restore a trashed file, revert a defaults write, or restore prior Finder tags / default app handler.

system_undo_listList the captured inverse operations available to undo, newest first.

system_uptimeGet system uptime since last boot.

system_url_openOpen a URL in the user's default browser or app.

system_volume_setSet the system output volume.

system_window_closeClose a window.

system_window_focusBring a window to the foreground and focus it.

system_window_fullscreen_setEnter or exit fullscreen for a window.

system_window_layoutProApply a multi-window layout preset (side_by_side, stack) across the visible windows.

system_window_listList visible windows on macOS, optionally filtered by bundleId, pid, or title substring.

system_window_minimize_setMinimize or restore a window.

system_window_moveMove a window to absolute screen coordinates (x, y) in pixels.

system_window_resizeResize a window to width × height in pixels.

system_window_snapProSnap a window to a screen region (left_half, right_half, top_half, bottom_half, top_left, top_right, bottom_left, bottom_right, center, maximize).

task_completeMark a task as completed.

task_createCreate a task to track a unit of work.

task_updateUpdate the status of an existing task.

workbench_createCreate a workbench of a registered kind in a project (when none is named, the project of the chat you are in, else the default project), seeded with the kind's content or with content you supply, which the kind checks.

workbench_deleteDelete a workbench and its Asset.

workbench_getRead one workbench: its kind, name, project, revision, content and view, plus what its kind reports beside them in `state` (an image or video workbench's undo and redo depth; for an image, which media its project does not carry).

workbench_listList workbenches with their id, kind, name, project and revision, optionally only one project's or one kind's, plus any workbench file that could not be read.

workbench_renameGive a workbench a new name.

workbench_update_contentReplace a workbench's content, its durable data, with an object the kind checks.

workbench_update_viewChange a workbench's view, its last-used working state (selection, zoom, layout).

Video

Automated video creation pipeline

image_to_videoStudioStudio video transform (image-to-video) — the kernel's headline 'animate this image' verb.

storyboard_add_shotStudioStudio video transform.

storyboard_assembleStudioStudio video transform.

storyboard_createStudioStudio video intent layer.

storyboard_generateStudioStudio video transform.

storyboard_getStudioReturns the current state of a video workbench's storyboard by id: its scenes, shots, each shot's kind / spec / clip reference / generationStatus, and stable ids.

storyboard_remove_shotStudioStudio video transform.

storyboard_reorderStudioStudio video transform.

storyboard_set_shotStudioStudio video transform.

storyboard_updateStudioStudio video transform.

text_to_videoStudioStudio video transform.

video_add_clipStudioStudio video timeline transform.

video_add_speechStudioStudio video timeline transform.

video_clipsStudioBrowse the PiperClips library (~/Documents/PiperClips/).

video_deleteStudioDelete a video screenplay and all associated artifacts (storyboards, recordings, narration).

video_edit_compositionStudioEdit visual/production settings for a screenplay.

video_edit_narrationStudioEdit narration and audio settings for a screenplay.

video_edit_regionStudioStudio video transform — apply a generative edit to just one TRACKED OBJECT across a whole clip.

video_edit_screenplayStudioEdit a screenplay's scenes and steps.

video_edit_timelineStudioEdit the post-recording timeline for a screenplay.

video_exportStudioExports a video workbench's timeline as an editorial interchange file for a pro NLE: OpenTimelineIO ('otio', default), Final Cut Pro XML ('fcpxml'), or CMX 3600 EDL ('edl').

video_export_timelineStudioExport a screenplay's edit timeline to an interchange format for use in external NLEs (Final Cut Pro, Premiere, DaVinci Resolve).

video_generate_speechStudioStudio video timeline transform.

video_getStudioGet a full video screenplay by ID with all scenes, steps, video config, narration config, and stage results.

video_import_mediaStudioImport an external video file (screen recording, stock footage, camera footage) into a screenplay's media library.

video_inspectStudioRenders composited frames of a video workbench's timeline back to the agent so it can SEE the current edit (the agent's eyes on the timeline): one frame at 'at_ms', 'count' frames from 'at_ms' to the end of the render, or 'count' (default 4) frames spread across it.

video_move_clipStudioStudio video timeline transform.

video_narrateStudioGenerate TTS narration audio for a screenplay's scenes.

video_previewStudioPreview a recorded screenplay.

video_recordStudioRecord a video by executing a screenplay — drives Chrome via CDP with human-like pacing while VisionPiper captures the screen.

video_rehearseStudioDry-run a screenplay via CDP.

video_remove_clipStudioStudio video timeline transform.

video_renderStudioPost-process a recorded screenplay: composite with frame styling, zoom, cursor overlays, color grade.

video_reorderStudioStudio video timeline transform.

video_saveStudioCreate or update a video screenplay.

video_segmentStudioStudio video transform — the region authority's video door.

video_set_clipStudioStudio video timeline transform.

video_set_keyframesStudioStudio video timeline transform.

video_set_render_rangeStudioStudio video timeline transform.

video_set_speechStudioStudio video timeline transform.

video_settingsStudioGet or update default video settings (preset, voice, language, pacing, capture, branding, subtitles, sound design).

video_splitStudioStudio video timeline transform.

video_timelineStudioReturns the current state of a video workbench's timeline: its tracks, each clip's stable id, source, in/out points, timeline position, speed, and any keyframes, plus the render range.

video_transcribeStudioStudio video timeline transform.

video_trimStudioStudio video timeline transform.

workbench_image_add_adjustment_layerStudioAdds an ADJUSTMENT layer to the TOP of an image workbench's layer stack: a layer with no pixels of its own that applies one classical image verb (levels, curves, black-and-white, a filter, …) to everything BELOW it, re-evaluated at every flatten.

workbench_image_add_content_layerStudioAdds an EDITABLE 'text' or 'fill' layer to an image workbench: a spec, not pixels, re-rendered at every flatten at its 'extent' (default the document size).

workbench_image_add_layerStudioAdds a raster layer to the TOP of an image workbench's layer stack from an existing `image` Asset.

workbench_image_crop_canvasStudioPhotoshop's Crop on an image workbench: reframes the document to a rectangle of its current box, given in document pixels ('x', 'y' from the top-left, 'width', 'height'; the rectangle must lie inside the document).

workbench_image_delete_layerStudioRemoves a layer (by its `lyr_...` id) from an image workbench.

workbench_image_duplicate_layerStudioCopies a layer in an image workbench and inserts the copy directly above the original, as one undoable step.

workbench_image_exportStudioWrites an image workbench's document to a real file the user can open elsewhere: 'png' (flattened, alpha kept), 'jpeg' (opaque, composited onto 'matte'), 'tiff' (alpha kept) or 'psd' (a LAYERED Photoshop document keeping every raster layer's pixels, mask, opacity, visibility and blend mode).

workbench_image_flattenStudioComposites every visible layer of an image workbench into ONE new immutable `image` Asset (canvas = the first raster layer's native size), applying each adjustment layer to the layers below it.

workbench_image_merge_downStudioMerges a layer onto the one directly below it in an image workbench: composites the two, mints the merged raster as a new `image` Asset, and replaces both layers with one layer referencing it.

workbench_image_reorder_layersStudioReorders the whole layer stack of an image workbench to match the given ordered list of layer ids (must be a permutation of the current ids; index 0 is the bottom / base).

workbench_image_resampleStudioPhotoshop's Image Size on an image workbench: scales the WHOLE document to new pixel dimensions.

workbench_image_set_canvasStudioPhotoshop's Canvas Size on an image workbench: gives the document its own pixel box and ground WITHOUT scaling any content.

workbench_image_set_layerStudioUpdates a layer in an image workbench — its blend mode, opacity, visibility, and name, and its CONTENT.

workbench_image_set_layer_maskStudioSets or clears a layer's matte in an image workbench.

workbench_image_set_layer_transformStudioUpdates a placeable layer's 2D placement in an image workbench: translation (document px, top-left origin), scale along the layer's own axes, rotation (degrees clockwise) and the pivot (layer px from its top-left, a pair).

workbench_image_set_layer_transformsStudioMoves, scales or rotates SEVERAL placeable layers in ONE step, whole or not at all, as one history entry.

workbench_redoStudioRe-applies edit(s) previously undone with workbench_undo on an image or video workbench, from the history its kind keeps: an image workbench's document or a video workbench's timeline.

workbench_undoStudioUndoes the last edit(s) to an image or video workbench, whoever made them (this agent, another agent, the person's editor, a voice session).

Video Tools

Video Tools platform: acquire, transform, serve, and capability introspection

media_dvrStudioBuffer a live HLS or DASH stream for local playback and rewind.

media_extractStudioDiscover the playable media on a web page and return candidate URLs (HLS/DASH manifests, progressive .mp4/.webm) using ONLY generic, standards-based signals: DOM `<video>`/`<source>`, Open Graph (`og:video`), Twitter player, JSON-LD `VideoObject`, oEmbed discovery, and — over the attached browser — live network manifests.

media_hostStudioHost a local video file OR a captured HLS/DASH manifest as a stable, range-capable artifact and return a bearer-free URL a browser `<video>.src` can play directly (no Authorization header needed — the unguessable `med_` id is the capability).

media_recordStudioRecord a live HLS or DASH stream from now.

stream_saveProSave a captured HLS (.m3u8) or DASH (.mpd) stream to disk using AVFoundation — downloads the segments at the chosen variant (no re-encode, no ffmpeg).

video_capabilitiesStudioList the native video layer's capability matrix: which codecs it can read, write, and mux (H.264 / HEVC / AAC are native, Opus / AC-3 re-encode the audio track, VP9 transcodes to HEVC, AV1 and VP8 are refused) and which containers it can read and write.

video_clipStudioTrim a `[startMs, endMs)` segment out of an already-hosted video artifact (by its `med_` id from media_host) into a flat, persistent .mp4 on disk via AVFoundation passthrough (no re-encode).

video_export_fileStudioSave an already-hosted video artifact to a real file in a user output folder — the save-to-disk / download step.

video_revoiceStudioReplace the voice in a video with generated speech, in one call, on this Mac.

video_upscaleStudioUpscale a video file 2x using PiperSR on Apple Neural Engine.

OAuth

OAuth connection management

oauth_connectStart an OAuth browser consent flow for a provider (Google, Reddit, GitHub, X).

oauth_connect_service_accountConnect a Google service account using a JSON key file.

oauth_disconnectDisconnect an OAuth provider.

oauth_statusCheck the authentication status of OAuth connections.

Looking for third-party servers to install?

This page lists ToolPiper's own built-in tools. The marketplace carries curated external MCP servers (GitHub, Slack, Postgres, and more), each installable in ToolPiper with one click.

Browse the MCP marketplace Back to the MCP server page →