More filters Product type Product type AI agent AI assistant Agent framework Automated workflow Coding agent Patent search tool Specialist AI tool Supervised agent Deployment Deployment A local native command-line application connected to a separately configured model backend. The tested distribution is official Fabric v1.4.507 Windows x86_64 at fixed release commit 32043e393cff877310ba02802d0e9e2a089f0b42. Model, API-server and optional web-interface deployment are separate choices. A self-hosted web application with installed translation models. Its documented native entry starts a local backend and the complete browser interface. Loopback binding and task-owned cache paths select a local evaluation configuration; they do not establish OS or whole-host network isolation. Recorded complete official GUI/backend: LibreTranslate 1.9.6, ArgosTranslateLT 1.12.1, CTranslate2 4.8.2 and MiniSBD 0.9.5 on the frozen CPU route. Two original GUI POSTs completed once each; the server was stopped after testing. Browser application with optional PWA installation or static self-host UI package. Chat storage is local by default; provider inference, configured retrieval/plugins and optional cloud sync have separate data paths. Static UI hosting or a PWA does not establish local or offline model inference. Cloud or self-hosted Cloud service Complete official LLM 0.36 CLI with a configured compatible model endpoint. Fresh LLM_USER_PATH selects application state and empty LLM_LOAD_PLUGINS set before import suppresses third-party discovery; two built-ins remain. These settings do not establish OS filesystem/network isolation. The two original native CLI cases were completed once each with LLM 0.36, Ollama 0.35.0 and the fixed existing Qwen2.5 Coder 1.5B Instruct Q4_K_M cache. Three runtime starts are retained: a pre-case ownership guard stop, primary completion followed by a before-boundary guard stop, and boundary-only completion. No original case was rerun for quality. Complete official Next.js application for local or self-hosted use. The reviewed main snapshot is commit 80d366a88cc8aa885a2d62ca9c60100eb72af40e with package version 2.1.1 and Node ^26.0.0, ^24.0.0 or ^22.0.0. The separately observed official release is v2.1.0; this snapshot is not presented as a new stable release. Application/browser storage, optional sharing database, inference backend and enabled external features have separate data paths. Complete official Node server and browser frontend for local or self-hosted use. The reviewed stable 1.19.0 tag maps to source commit 7e8663cd9c184a550b37238218bdd32c6efc68e9; the source requires Node >=20. Application data is stored in its configured dataRoot. The frontend, connected inference backend and optional extensions have separate network and data boundaries. Desktop or CLI with hosted model and cloud-agent options Google-hosted verification portal; no embedded uAgentKit scanner or native store connection is provided Hosted Chatbase workspace and customer-facing widget/page; connected channels, API and actions require account entitlements and customer-managed permissions. Hosted Postiz Cloud or self-hosted stack with database, Redis and Temporal; social-provider configuration varies by deployment. Hosted Typefully workspace, with a web app, Mac app and client-side MCP or Agent Skill connections; self-hosting not verified. Hosted browser resume workspace with a documented conversational AI Resume Agent. On 2 October 2026, clicking its public Try for free CTA opened app.rezi.ai/signup with Google, Apple and email/password account controls and no visible guest continuation. Legal separately describes no-account use; this one observed entry does not establish that every guest path is absent. No prompt, upload, login, registration, generation or export was performed; actual Agent Free entitlement and quota remain unknown. Hosted browser workspace: save an uploaded product, choose Custom or Templates, generate on the service and download drafts from Recent. Source-image upload, hosted generation, credit usage and export are separate observable steps; no marketplace publication or selling-system action is established. Hosted image/video application. Reference-image upload and a downloadable asset do not establish local processing or a native commerce connector. Hosted note-taking service with a web entry and documented desktop/mobile apps. Recordings and notes use the service’s cloud paths; microphone permissions, connected integrations and sharing are separate. A desktop installation does not establish local or offline inference. Hosted notes and cloud rewriting, with optional on-device transcription in supported native apps. Web and Chrome-extension transcription always use the cloud; offline transcription does not make the whole service offline or self-hosted. Hosted persistent cloud computer with desktop and mobile clients; official docs list macOS, Windows, Linux, iOS and Android. Hosted personal workspace. Exact hosting regions, enterprise tenancy and self-hosting were not established in the reviewed sources. Hosted service; exact tenancy and regional options require confirmation Hosted web service; region, tenancy and account data controls require vendor confirmation. Hosted web tools with local browser storage for the free experience; optional Pro stores account content on the operator’s servers. Hosted web workflow and channel integrations; local video upload is an input method, not local model execution. IDE or CLI assistance with hosted and GitHub cloud-agent services Local Desktop on macOS, Windows and Linux; self-hosted server/Docker deployment; or optional vendor Cloud hosting. Local privacy depends on chat/embedding providers, tools, web access and telemetry configuration. Local desktop, IDE, CLI, or embedded runtime with chosen model endpoint Local library runtime; optional managed deployment and observability services Local workspace with an official desktop application, mobile distributions and a browser-accessible self-hosted kernel. The reviewed pilot preparation pins release 3.8.6 to source commit 158812497497c6fa8a6d1031bfa2f025a61ee5c2; actual running version and package identity must be retained separately. Hosted access, cloud sync and the chosen AI endpoint introduce their own configuration and data boundaries. Locally operated framework or separately licensed managed/enterprise platform Managed cloud or documented local/self-hosted server; local execution requires a configured model provider, browser and database. Managed cloud, private/VPC, or self-hosted Community Edition Meta-hosted personal cloud computer; mobile, Mac and WhatsApp access. No self-hosted distribution verified. Mobile app Official Community Edition source with desktop and complete standalone web renderer; the web platform uses local browser storage. Local provider processing and all-app network activity must be checked separately. Official Mods v1.8.1 Windows x86_64 release ZIP matches official checksums; binary digest retained. Fixed commit 9b36cf882989cf6807ec82f8afbc56d418ba9264. Native dedicated XDG mods.yml selects one local OpenAI-compatible API, native JSON mode, temperature 0, top_p 1 and max_tokens 800. No-cache, max-retries 0, no MCP servers or model fallback; product source unchanged. Official Pi Windows x64 1.0.2 ZIP SHA 37a2ee3c5222a73efc78b49f571748f11a74f4dbe160da3030a404da71f464e7 matches GitHub asset digest; fixed commit cd32f7725fdbddbaecdff5b1e68491563394e0ca. Unpatched pi.exe --print --mode json with synthetic stdin, frozen custom system instruction and separate PI_CODING_AGENT_DIR. Offline startup, no tools/extensions/skills/prompt templates/themes/context files/session persistence; retries/compaction off. No project-resource discovery; --no-approve is not a sandbox. Official PyPI 0.34.0 wheel SHA-256 cc33e2dac71def4660ade7a89d2d9f7cc0a05df85d16036fcc5e07e13ce5a407 verified; fixed release commit a401cd1f29aa48115b92c1b5410caa9ed485fb78. All 483 installed product Python modules match the wheel. Full wheel-to-release-source correspondence unverified; source unchanged. Python 3.12.14 dedicated environment. Native noninteractive chat; tools none, no workspace/context_cmd, no prune summary model, no stream, original full-noexamples prompt. Injection hygiene explicitly off; this pilot tests text drafting, not file/shell tools. Official PyPI 0.86.0 wheel SHA-256 386be777c7197494a020337f18c15660fe32a5b4a350123e583bd07ad25365d8 verified; fixed release commit a4be6ccd87ebaa59b361f3f028d116ce1761b626. All 81 installed product Python modules match the wheel. Full wheel-to-release-source correspondence unverified; source unchanged. Python 3.12.14 dedicated environment. Native whole-file editing; no git, auto-commit, lint, test, analytics, updates, repo map, restore or shell suggestions. Any new-file confirmation receives n. Known status.md can be edited without confirmation; this is not an OS sandbox. Official PyPI Chatblade 0.7.0 source distribution matches its published SHA-256; fixed commit 9f12a3405a1daf0f09da0e0889d97ac366ceaa7b. All 10 installed Python module files match the source. Python 3.12.14 task-only venv with OpenAI SDK 1.35.15 and httpx 0.27.2 compatibility constraint. Native custom --openai-base-url, --extract --only --no-format, temperature 0 and non-streaming; product source unchanged. Native scratch sessions are saved only in the dedicated lab profile. Official PyPI shell_gpt 1.5.1 wheel, SHA-256 matched to distribution metadata, fixed source commit dee88ff87bb93899971a3ca1361ad74678e4a94f. All 20 installed sgpt Python files matched the source. Python 3.13 task-only venv used OpenAI SDK 2.54.0, Typer 0.27.2 and prompt_toolkit 3.0.53. A missing click import was repaired before cases by installing click 8.5.0; product source was unchanged. Dedicated subprocess profile, roles and caches kept configuration in the lab. Official Rust terminal distribution with local product configuration and session state. Current fixed source uses INTERPRETER_HOME for a task-specific existing state directory and deliberately ignores CODEX_HOME; that state setting does not prove isolation of shared skills or all discovery roots. On Windows, native sandboxing requires explicit configuration: unelevated uses RestrictedToken and full disk read, so it does not establish strict denied-read isolation outside a workspace. Interactive approval controls and headless exec defaults are separate. Record the actual request protocol; a Chat flag alone does not prove Chat transport in the OSS exec path. Official Windows amd64 v0.9.9 ZIP SHA-256 c177435d00152df1e6c52bf3909a7ced78a7bb2eebecadfe79ac93fd816e066f matches release checksums. Fixed source commit a1c16f4763340bbebf23fe6b50be519019224e21; native version v0.9.9+a1c16f47. No source patch; full binary reproducibility unverified. One no-tools natural-language .gpt program, native stdin, separate task config/file credential store, --confirm, no cache/retries/UI/chat; cached local model, nonstreaming. Official Windows x64 CLI 0.97.1 ZIP SHA 168b5c40caee6242acff82c02b7239ad6bdb02291e253eb03855fa4c7e063bec matches GitHub asset digest and checksums. Fixed commit e0baf255be53f932042b377630e233d4c33c3b4b. Native in-process crush run --quiet with stdin and instruction argument, task-only config/data/cache/skills and separate case data directories. All 29 native tools disabled; no MCP/hooks/LSP, metrics/discovery/provider updates/summarization, no yolo/permission bypass. Default coding system/context prompt preserved; title calls separately audited. Official Windows x64 CLI 1.53.0 ZIP SHA 3a951c661f12415f7947daac2bb4651af1a7b41532f34d7c2f05ea6636eadaa9 matches GitHub asset digest. Fixed source commit 76da81cb964b21cd096db739302329b40c2998b8. Unpatched native goose run with stdin, --no-profile, --no-session, quiet text, explicit openai provider and cached model, additional system instruction and max-turns 2. Separate native GOOSE_PATH_ROOT, no requested extension/builtin and telemetry off. Full binary reproducibility and broader native tool workflows unverified. Official Windows x64 baseline v1.18.34, ZIP SHA f89ab2720050780a450e3cf3e48ac3f0409235b46b6c548c69aa2b7051d716f4 matches GitHub asset digest. Fixed commit aec0b9a6d8898f68f923aaf08b7306d931fd9d76. Unpatched native CLI; custom fixture primary Agent, global/agent deny permissions, model tool_call=false, disabled built-in agents, explicit title and separate task XDG profile. No share, MCP, plugin, external skills, LSP, formatter, project config, updates or compaction. Frozen steps=1 confounds task output on the first call. Official Windows x86_64 AIChat 0.30.0 ZIP, verified against the release asset SHA-256, at source commit 430416d914896c3534c04b84c0226910c64e3e66. A dedicated AICHAT_CONFIG_DIR selected the existing llama.cpp OpenAI-compatible endpoint and cached Qwen model. No container, database or new model weights were needed for CMD-mode drafting. Official Windows x86_64 ZIP 0.3.1 digest c76e4b9f3f95137b8cb58229e756538b12ed0fee0e4301a01b531626bf740af1 matches release; fixed commit 2cf030d2fd3b871d7ec17e3be34c24688aac76da. Unmodified CLI agent -m, dedicated PICOCLAW_HOME/CONFIG, native turn profile history/skills/tools off, default system context retained. No gateway/launcher/channel/MCP. Native retries remain 2, not disabled. Official hosted browser webapp. Reviewed source routes an ordinary local file through FileReader into a browser worker; page initialization may load processing resources. Self-hosting, complete offline operation, current model identity and network isolation have not been established. Official nanobot-ai 0.3.5 Windows x64 wheel SHA 60134447b04bfb290938f7abcbfc4a373ac5daffe3ac58f9ea784a2dc2d25a6e matches PyPI digest; fixed commit 1bb712d3488915ca4ed9ccc1a93067ff722f5ab9. All 329 installed Python modules match both official wheel and fixed source. Native AgentLoop.from_config + ToolRegistry.unregister + process_direct with separate synthetic workspaces and ephemeral turns; all actual tools removed. Stock CLI only version/help, no gateway/TUI/scheduler/channels or MCP. Default native system/context retained; installed skills disabled. Official native web UI. Three synthetic pasted-text sources selected in one test-only notebook; two once-submitted original chat prompts. Observed UI label Gemini Notebook. Exact vendor model version, inference runtime, token usage and hidden backend request counts unobserved. Official tgpt-amd64.exe v2.15.0 verified against its GitHub release digest, fixed commit 21b3778945f08c81f9d666477035444290738567. Native --whole --quiet with explicit --provider openai and full loopback chat-completions URL; empty dedicated config and profile. No source changes, fallback providers, tools, MCP, shell execution, search or business account. OpenAI-hosted cloud computer; creation through desktop web or the ChatGPT desktop app. Mobile-app conversation follows setup; mobile web is unsupported in the reviewed help article. Pin the complete official rembg 2.0.85 package with cpu and cli extras in a task-owned Python environment. Native rembg s exposes a Gradio UI and HTTP API; bind it to 127.0.0.1 on a task-owned port. The frozen inference cases use the native multipart HTTP route, not UI interaction, a direct library wrapper or isolated ONNX inference. Both REMBG_HOME and U2NET_HOME point to the same task-owned model root so the latter precedence and legacy lookup cannot reuse an unrelated user model. Verify actual application/cache state separately; these settings do not establish an OS sandbox or network isolation. Recorded complete Agent Canvas Docker sandbox; hosted and other CLI routes have separate prerequisites Self-hosted browser application installed with a supported Python runtime or pinned official container, connected separately to local Ollama or an authorized model endpoint. Hosting the interface does not supply an LLM or prove an offline deployment. The public community website is distinct from the locally deployed application. Subtitle Edit 5 is documented as a cross-platform desktop application. SeConv is the official headless utility that reuses libse/libuilogic without a GUI dependency. The recorded component was the complete SeConv-Windows-x64.zip v5.2.0 release asset at source commit d8e3b8b41e856a896c541ce7e59b490a99c21196. Its official build publishes a self-contained Windows distribution and copies the MIT license. The complete ZIP, extracted native executable, distribution inventory, original fixtures/prompt, existing model and runtime were bound in the pre-inference manifest. Each official native process exited 0 and saved its first SRT; both cases nevertheless failed their original semantic conditions. These component/runtime checks do not establish desktop GUI or full product-quality acceptance. The editor is opened at photopea.com in a web browser; the official introduction recommends a large screen, precise pointing device and keyboard for comfort. Official ordinary-editing documentation says opened files are processed on the user's device. The embedding API separately describes optional third-party Dezgo configuration for Magic Replace and Remove BG, including a key and request URL. The current native Remove BG route, provider, network processing, model, credits and build must be observed separately and are unknown here. The 22 April 2024 Photopea 5.6 post is a dated release reference, not the recorded current native version. The official complete Windows Vibe 3.2.2 distribution was downloaded/hash-verified and extracted in a task-owned directory without invoking the installer. It includes vibe.exe, native vibe-server.exe and ffmpeg.exe. Actual native server --version identified v0.6.10. Two original plain-text native transcribe cases each ran once using its bundled server and one pinned local Whisper tiny model. Native processes completed and task-owned app/server/FFmpeg process checks found no leftovers. The desktop GUI and local HTTP API were not invoked; actual inner CPU/GPU backend and whole-host network behavior were not measured. The official online UI is voice-generator.pages.dev. Select Execution place Browser for browser-local inference; API mode is a separate self-hosted branch. The frozen cases use CPU and the model_quantized 8-bit option, so the planned route does not require WebGPU. First generation loads the fixed ONNX model and voice plus eSpeak NG, ONNX Runtime and FFmpeg browser resources. The native UI displayed v0.1.3; its bundle was not bound to the retained source commit. The public repository snapshot was last pushed on 16 March 2025. The recorded route uses the complete fixed official source package, metadata version 0.16.0 at commit 0957c301ed06c2a5857b834358c7227c739041d4, installed with its official requirements in an isolated Python 3.13 environment. Its native library calls an existing local OpenAI-compatible llama.cpp backend. The separate web GUI and hosted service were not tested. The recorded self-hosted pilot used the complete official source commit ae229ca894c0b80ad84664afcfdde523b5e87057, Python 3.12.14, PostgreSQL 16.2 with pgvector 0.6.2, fixed local retrieval models and Qwen2.5 Coder 1.5B Instruct Q4_K_M through Ollama 0.35.0. Its original web interface was built; native file-index and chat endpoints were exercised on localhost. The PostgreSQL wheel required a disclosed timezone-data supplement; Khoj source and original requests were unchanged. This task-owned application configuration is not OS isolation, and GUI interactions were not evaluated. Use the full official web application at scribeocr.com. Official documentation also supports serving the complete repository through a local HTTP server; no desktop installation is provided in the retained README. Built-in recognition is described as browser-local, but code, WASM and language resources may load over the network. Optional AWS Textract sends recognition to its separate cloud route. The retained GUI source commit and Scribe.js gitlink are research references, not verified identities of the current hosted build or downloaded traineddata. Vendor-hosted web application Vendor-hosted web service; backend model and inference runtime are not independently observed. Language Language Not verified API API Yes — verified Not verified Self-hosting Self-hosting Yes — verified Not verified Open source Open source Yes — verified Not verified Verification status Verification status partially verified Platform Platform Amazon Bluesky Facebook Instagram LinkedIn Mastodon Mercado Libre Ozon Shopify Threads TikTok TikTok Shop Website X YouTube Platform relationship Platform relationship Content output Marketplace data Native integration Integration method Integration method .gpt natural-language program ACP AGENTS.md and shared skills AI Resume Agent API API/SDK Advertised local HTTP API Agent Skills App connectors Authorized business connectors Authorized data connectors Bring your own model provider Bring your own translation backend Browser CPU and optional WebGPU choices Browser UI with native audio playback and Download Browser access to a self-hosted kernel Browser editor Browser embedding and configuration API Browser interface Built-in CLI CRM integrations Chrome extension Cloud browser Compatible model provider Complete browser UI with native document import and Download Configurable model and retriever providers Configurable primary Agent Configurable provider Configured OpenAI-compatible endpoint Configured OpenAI-compatible model endpoint Configured enterprise integrations Configured messaging channels Custom OpenAI-compatible provider Custom or template generation Desktop app Desktop subtitle editor Desktop/mobile apps Docker Documented REST API Documented provider, role and file-input support Documented stdin, shell, code, chat and function modes Documented web interface Downloadable media Enterprise connectors File export Framework GitHub integration Helpdesk integrations IDE extension Image download Job-description keyword analysis LSP and MCP (not exercised) Linux desktop app Local Gradio web UI Local Ollama configuration Local Ollama connection Local browser app Local command line Local or cloud model providers Local terminal CLI Local translation API MCP MCP extensions (not exercised) MCP with OAuth Mac app Messaging channels Mobile app Mobile apps Model-provider connection Native Ask AI Native CLI, TUI and WebUI (CLI preflight only) Native HTTP image-upload API Native Python AgentLoop SDK Native Python CLI on Windows/macOS/Linux Native Windows CLI Native Windows non-interactive CLI Native Windows/macOS/Linux CLI Native bundled-server CLI Native chat API Native command line with a configured compatible model endpoint Native configurable channels (not exercised) Native exec Native file-index API Native single-message CLI Native terminal CLI Native voice recording Official Python library Official headless SeConv CLI Official native CLI Official web UI OpenAI-compatible local endpoint OpenAI-compatible local provider OpenAI-compatible model endpoints OpenAI-compatible provider Optional Claude API or Ollama analysis Optional MCP Optional PWA Optional cloud storage access Optional configured external tools Optional configured integrations Optional local HTTP hosting of the complete official application Optional plugins Optional self-hosted OpenAI-compatible TTS API PSD and image file editing Plan-dependent PDF export Print/JSON, RPC and SDK modes Product image upload Python library REST API Reference-image upload Resume editor Salesforce integration Scheduled routines Self-hosted Self-hosted Next.js server Self-hosted Node server Self-hosted application Self-hosted server Self-hosted web app Separate optional AWS Textract cloud adapter requiring user credentials Shopify integration Slack Slack integration Social scheduling connections Standalone agent page Standalone browser verification portal Static self-host package Stdin text transformation Third-party automation connectors Video upload Web app Web widget Webhooks WhatsApp Windows desktop app Word add-in YouTube source import Zendesk integration Zotero integration macOS app macOS desktop app
uAgentKit
