Releases: joliver1981/everiapp
Release list
v0.17.3
EveriApp v0.17.3
Your apps can now run Python on the server. This release brings a real
server-side compute layer to EveriApp — so the apps you build with a sentence
can crunch data, generate spreadsheets and PDFs, chain AI steps, and use the
Python ecosystem, all without leaving the browser-based builder. By EveriAI LLC.
Windows only for now. Linux servers can run deployed apps via the EveriApp Agent.
Downloads
| Download | What it is | Who needs it |
|---|---|---|
| EveriApp-Setup-win-x64-0.17.3.exe | The EveriApp Builder — the main desktop application | Everyone |
| EveriApp-Agent-Setup-win-x64-0.1.0.exe | The EveriApp Agent — a small service for deploying apps to a server | Only if you deploy apps to remote machines |
The .sha256 files are checksums for verifying your download (see below).
What's new
🐍 Server functions — Python that runs on the platform
Generated apps used to be browser-only. Now an app can include server
functions — Python that runs on the EveriApp host and returns results to the
app's UI. Ask the builder for something the browser shouldn't do, and it wires
one up for you:
- Crunch real data — aggregate thousands of rows with pandas/numpy
and return just the answer, instead of shipping everything to the browser. - Generate documents — real .xlsx workbooks (openpyxl) and PDFs
(reportlab / pypdf), downloaded straight from the app. - Orchestrate — query the app's data, call an external API, run an AI step,
and write back — server-side, in one call. - Keep secrets off the client — API keys and logic stay on the platform.
Each function runs in an isolated Python process with a hard timeout. The
following libraries are bundled and ready to import: pandas, numpy, openpyxl,
reportlab, pypdf, python-dateutil (plus the full Python standard library).
📦 Admin → Python Packages
A new admin page lets you install additional Python packages from PyPI for
your server functions to use — with live version lookup, one-click
install/uninstall, and a configurable package index URL for air-gapped or
internal-mirror (Artifactory / Nexus) deployments. The installer now bundles a
CPython runtime and pip, so this works on a fresh machine with no separate
Python install required.
🔌 Smarter data & connections
- Apps discover their attached connections at runtime, so a newly attached
connection appears without regenerating the app. - The builder is now honest about your databases: instead of inventing table
names and filling the screen with fake sample data, it guides you to create a
Dataset over your SQL connection and uses your real columns — no more
plausible-looking-but-wrong queries hiding behind demo data.
🛠️ Reliability & quality
- Fixed platform-wide errors during AI generation that could surface as
intermittent failures when the builder was used under load. - Generated demo data is now seeded reliably — no duplicated rows on first
load, no partial seeds from rate limits.
Upgrading from an earlier version
Run the new installer over your existing install — it upgrades in place and
your data under C:\ProgramData\EveriApp is preserved (apps, users,
connections, settings). The EveriApp Windows service restarts on the new
version automatically. Server functions and the Python Packages page are new
capabilities available immediately after upgrading.
Install the Builder
- Download EveriApp-Setup-win-x64-0.17.3.exe and run it (requires admin
rights — it installs a Windows service).- Windows SmartScreen may warn that the publisher is unrecognized — this
release isn't code-signed yet. Click More info → Run anyway, and verify
the download with the checksum below if you'd like.
- Windows SmartScreen may warn that the publisher is unrecognized — this
- Choose a port when prompted (default 8800).
- The installer starts the EveriApp service and opens your browser.
- Create (or sign in to) your administrator account.
- Open AI Providers, add a provider + API key, and start building.
Installs to C:\Program Files\EveriApp, runs as the EveriApp Windows
service, and stores data under C:\ProgramData\EveriApp.
Install the Agent (optional — only for remote deployments)
On each server you deploy to, run EveriApp-Agent-Setup-win-x64-0.1.0.exe,
paste the agent token (create one in EveriApp under Secrets, category
agent_token), then add a Deployment Target in EveriApp pointing at that
server's IP on port 8765.
v0.13.1
EveriApp 0.13.1
This release makes AI providers a first-class part of the platform: connect OpenAI, Anthropic, OpenRouter, or Azure OpenAI in a couple of clicks, curate which models your apps can use, and let the AI builder wire them straight into the apps it generates.
Rolls up all changes since 0.7.12.
🤖 AI Provider Connections — connect a provider in two clicks
- New "AI Provider" connection type in Admin → Connections, with built-in presets for OpenAI, Anthropic, OpenRouter, and Azure OpenAI — plus a Custom option for any OpenAI-compatible endpoint (vLLM, Ollama, LM Studio, gateways).
- No more hunting for base URLs or auth headers. Pick a provider and everything is prefilled — just point it at the secret holding your API key.
- Live model lists. Click Fetch models and EveriApp pulls the current model list straight from the provider. Curate which models apps may offer, and set a default model.
- Test that actually tests. Test connection now validates your API key against the provider — a bad key fails the test with a clear message instead of passing silently.
- Secrets dropdown. Wherever a connection needs a credential, you now pick the secret from a dropdown instead of retyping its name — and dangling references (deleted/renamed secrets) are flagged visibly.
⚡ Apps can now call real external APIs and LLMs
- Generated apps can make real, server-brokered calls to any admin-approved API or AI provider. API keys never leave the server and are never embedded in app code.
- Access is opt-in per app: mark a connection app-callable, attach it to the app, done. Calls are rate-limited and audit-logged.
- Apps discover their connections at runtime — attach a new provider and it appears in the app instantly, no rebuild.
- One request shape for every AI provider: apps built on your connections get chat, model pickers, and true side-by-side model comparisons across vendors.
🧠 A smarter app builder
- The AI builder now understands what the platform can and can't do. Ask for an integration that isn't configured yet and it tells you exactly what to set up — instead of inventing a fake, simulated stand-in.
- Attached data sources, connections, and providers are wired in zero-config: the app just works the moment resources are attached, with no "paste your API key" screens.
- Longer AI outputs: output limits for in-app AI decisions and the app assistant were raised substantially, and arbitrary input-size caps were removed.
🛡️ Reliability & safety
- The AI bug-report analyzer can no longer propose changes to platform-owned files — fixes stay scoped to your app's own code.
- Creating a new app is now near-instant (dependencies are provisioned in the background on first preview).
⬆️ Upgrading
Download EveriApp-Setup-win-x64-0.13.1.exe and run it over your existing installation — it upgrades in place and restarts the EveriApp service automatically. Windows x64 only.
SHA-256
v0.7.4
EveriApp Builder v0.7.4
This release turns the builder from "generate an app" into generate, observe, and fix an app — a full runtime tracing spine, an AI copilot that watches you test, mini-LLM decisions as first-class resources, and in-chat mockups. Plus a wave of reliability fixes to the live preview.
On-prem, single-tenant, Apache-2.0. No data migration required — schema columns are added idempotently on startup.
✨ Highlights
🔎 App Inspector — see everything your app does
A new Inspector panel sits beside the Preview and narrates the running app in plain English as you click through it:
- Story mode — "You clicked Save order → the app asked the database → the database refused: customer_id can't be empty" — a chronological, human-readable trace of clicks, data calls, AI calls, and errors.
- Timeline mode — the same events as a developer waterfall with status, latency, tokens, and cost.
- Live follow — streams events in near-real-time while you test.
- Routine successful data calls collapse into one expandable line so the story doesn't drown in plumbing.
- Sessions are grouped by trace ID; one clear-trace button gives you a fresh slate.
🤖 Test with AI — a copilot that debugs alongside you
The Inspector surfaces issues and helps you fix them without leaving the flow:
- Detect (free, rules-only) — repeated errors, swallowed errors, decisions falling back, and slow calls appear as chips.
- Explain — click an issue (or an error row) and AI diagnoses it from the trace and your app's real source, naming the root cause and the files involved.
- Fix it — one click turns the diagnosis into a fix request that runs through the normal build pipeline (self-heal + verification + hot reload) so you watch it happen.
🧠 aiDecide — mini-LLM decisions as first-class resources
Build genuinely agentic apps without brittle regex. Declare a named decision (aiDecide('classify_question', …)) and the platform owns the prompt, schema, model, cache, and fallback:
- Prompt-as-data: tune a decision's prompt in the platform and the next call uses it — no rebuild.
- Never crashes an app: timeouts, provider errors, and off-schema answers resolve to a declared fallback.
- Born observable: every decision is traced, cost-metered, cached (user-scoped), and rate-limited.
- The generator now reaches for decisions instead of keyword/regex heuristics for fuzzy logic.
🎨 Mockups & diagrams in the builder chat
Plan visually before building:
```mermaidblocks render as flowcharts, sequence, and ER diagrams.```mockupblocks render as sandboxed screen mockups — with a "Make this real" button that implements the design as actual screens.
📚 Developer skills — standing preferences the AI follows
- Personal skills (sidebar) — "always enable WAL on SQLite," "every list gets a loading state" — applied to every app you build.
- Org conventions (Admin → Platform) — house standards applied to every developer.
🎛️ Per-purpose AI providers + full cost visibility
- Assign a default provider (and model) per purpose — pin a fast, cheap model to app decisions and a stronger one to generation.
- Every LLM call path is now cost-metered (in-app assistant, self-heal, bug analysis, decisions, copilot) — previously several were invisible.
🐛 Bug reports now carry the trace
Submitted bug reports include the session's traced events, so the AI triage sees what happened chronologically instead of guessing from a console tail.
🚀 Improvements
- Version indicator in the sidebar shows UI vs API version and flags a mismatch — no more debugging a stale bundle or backend.
- Decision robustness — per-decision timeouts (default 30s, up to 120s), a larger output ceiling for generative decisions, and self-documenting timeout/truncation errors that name the exact knob to change.
- Decision registry self-heals from the app's manifest on every preview start, with a manual re-sync endpoint.
- Action-oriented tooltips across the builder — each control states what it does and what clicking changes.
- Marketplace publish prefills listing metadata (short description, category, tags, license) so it stays consistent across publishes.
🔧 Fixes
- Live preview reliability — fixed blank preview at startup (readiness handshake, WebSocket relay, heartbeat), fixed resets during a session (transparent WS relay; rotating token no longer in the iframe URL), and fixed a 401 regression from a stale auth-token selector.
- Setup wizard hardening — invalid or malformed wizards are rejected on every write path instead of being stored broken and 500ing later; garbage rows degrade gracefully.
- App viewer loads previews through the runtime proxy with a scoped cookie, keeping auth tokens out of URLs and logs.
- Vendored SDK is refreshed on package import and marketplace install, so already-fixed SDK bugs don't reappear in imported apps.
Full details: see the commit history from the initial public release through v0.7.4.
v0.1.0
EveriApp v0.1.0
The first public release of EveriApp — build AI-powered apps from a
natural-language description, then share and deploy them. By EveriAI LLC.
This is an early release for Windows. macOS and Linux are on the roadmap; Linux
servers can run apps via the EveriApp Agent today.
Downloads
| Download | What it is | Who needs it |
|---|---|---|
| EveriApp-Setup-win-x64-0.1.0.exe | The EveriApp Builder — the main desktop application | Everyone |
| EveriApp-Agent-Setup-win-x64-0.1.0.exe | The EveriApp Agent — a small service for deploying apps to a server | Only if you deploy apps to remote machines |
The .sha256 files are checksums for verifying your download (see below).
What's in this release
The Builder
- Describe an app in plain language and the AI generates a working React app —
with live preview, a built-in code editor, and immutable versioning + rollback. - Create your own admin account on first launch (username + password) — no
default credentials to look up. - Add local users and reset their passwords from Users & Roles; optional
single sign-on via LDAP / Active Directory, SAML, or OpenID Connect. - Bring your own AI provider (Anthropic, OpenAI, and more) in AI Providers.
- Marketplace — browse and install apps from this server or the public
EveriApp Marketplace, and one-click publish your own apps (screenshots
captured automatically). - Connect external databases (Connections & Datasets), with SQLite built in.
- Security scanning of generated code, dependency scanning, and an optional
publish-approval workflow. - Node.js is bundled — nothing else to install.
The Agent
- A lightweight Windows service you install on each server you want to deploy to.
- Receives app versions from the Builder, runs them, and reports health back.
Install the Builder
- Download EveriApp-Setup-win-x64-0.1.0.exe and run it (requires admin
rights — it installs a Windows service).- Windows SmartScreen may warn that the publisher is unrecognized — this
release isn't code-signed yet. Click More info → Run anyway, and verify
the download with the checksum below if you'd like.
- Windows SmartScreen may warn that the publisher is unrecognized — this
- Choose a port when prompted (default 8800).
- The installer starts the EveriApp service and opens your browser.
- Create your administrator account.
- Open AI Providers, add a provider + API key, and start building.
It installs to C:\Program Files\EveriApp, runs as the EveriApp Windows
service, and stores data under C:\ProgramData\EveriApp.
Install the Agent (optional — only for remote deployments)
- On the target server, run EveriApp-Agent-Setup-win-x64-0.1.0.exe.
- Paste the agent token when prompted (create one in EveriApp under
Secrets, categoryagent_token). - Back in EveriApp, add a Deployment Target pointing at that server's IP on
port 8765.
System requirements
- Windows 10 or 11 (64-bit)
- ~2 GB free disk space
- 4 GB RAM minimum, 8 GB recommended
- An API key from an LLM provider (e.g. Anthropic or OpenAI) to use the builder
Verify your download (optional)
In PowerShell, in the folder where you downloaded the file:
(Get-FileHash .\EveriApp-Setup-win-x64-0.1.0.exe -Algorithm SHA256).HashIt should match the value in EveriApp-Setup-win-x64-0.1.0.exe.sha256:
b3818f8772e595f5b63c331f906e79c148397edb04cac4d7d1ae5008ca9c0889 EveriApp-Setup-win-x64-0.1.0.exe
2532932a5895e7ad37c9bad4622b21e5d111ada06b801d2477be295b673dafdc EveriApp-Agent-Setup-win-x64-0.1.0.exe
Known limitations
- Windows only for now — macOS and Linux desktop builds are planned.
- Not yet code-signed, so Windows SmartScreen shows an "unrecognized
publisher" prompt on first run. Verify with the SHA-256 checksum above. - Early release — expect rough edges, and please share feedback.
© EveriAI LLC
v0.17.4
EveriApp v0.17.4
Your apps can now run Python on the server. This release brings a real
server-side compute layer to EveriApp — so the apps you build with a sentence
can crunch data, generate spreadsheets and PDFs, chain AI steps, and use the
Python ecosystem, all without leaving the browser-based builder. By EveriAI LLC.
Windows only for now. Linux servers can run deployed apps via the EveriApp Agent.
Downloads
| Download | What it is | Who needs it |
|---|---|---|
| EveriApp-Setup-win-x64-0.17.4.exe | The EveriApp Builder — the main desktop application | Everyone |
| EveriApp-Agent-Setup-win-x64-0.1.1.exe | The EveriApp Agent — a small service for deploying apps to a server (versioned separately) | Only if you deploy apps to remote machines |
The .sha256 files are checksums for verifying your download (see below).
What's new
🐍 Server functions — Python that runs on the platform
Generated apps used to be browser-only. Now an app can include server
functions — Python that runs on the EveriApp host and returns results to the
app's UI. Ask the builder for something the browser shouldn't do, and it wires
one up for you:
- Crunch real data — aggregate thousands of rows with pandas/numpy
and return just the answer, instead of shipping everything to the browser. - Generate documents — real .xlsx workbooks (openpyxl) and PDFs
(reportlab / pypdf), downloaded straight from the app. - Orchestrate — query the app's data, call an external API, run an AI step,
and write back — server-side, in one call. - Keep secrets off the client — API keys and logic stay on the platform.
Each function runs in an isolated Python process with a hard timeout. The
following libraries are bundled and ready to import: pandas, numpy, openpyxl,
reportlab, pypdf, python-dateutil (plus the full Python standard library).
📦 Admin → Python Packages
A new admin page lets you install additional Python packages from PyPI for
your server functions to use — with live version lookup, one-click
install/uninstall, and a configurable package index URL for air-gapped or
internal-mirror (Artifactory / Nexus) deployments. The installer now bundles a
CPython runtime and pip, so this works on a fresh machine with no separate
Python install required.
🔌 Smarter data & connections
- Apps discover their attached connections at runtime, so a newly attached
connection appears without regenerating the app. - The builder is now honest about your databases: instead of inventing table
names and filling the screen with fake sample data, it guides you to create a
Dataset over your SQL connection and uses your real columns — no more
plausible-looking-but-wrong queries hiding behind demo data.
🛠️ Reliability & quality
- LDAP directory test & search now use your configured identity provider.
Previously the Users & Roles page's "Test connection" and user search ran
against a development stub and answered "Mock mode" even with LDAP fully
configured. They now test and search the LDAP provider you set up under
Admin → Platform → Authentication, using its new optional service account
DN — and failures explain themselves (bad service bind, unreachable
server, provider not configured) instead of returning nothing. - Fixed OpenAI providers failing with "LLM Provider NOT provided" for
current model names (including the setup page's own defaults). Model routing
is now always explicit, so brand-new OpenAI models work the moment they
ship — no platform update required. - Fixed a black empty page after install on servers whose Windows registry
maps.jsto a non-JavaScript file type (browsers then refuse the app
bundle). Asset MIME types are now pinned and no longer depend on the host
machine's registry. - Fixed a shutdown hang that could make service stops and restarts slow or
unclean — the platform now winds down its background work in order and
exits promptly. - Fixed platform-wide errors during AI generation that could surface as
intermittent failures when the builder was used under load. - Generated demo data is now seeded reliably — no duplicated rows on first
load, no partial seeds from rate limits. - The browser tab now shows a proper EveriApp icon.
Upgrading from an earlier version
Run the new installer over your existing install — it upgrades in place and
your data under C:\ProgramData\EveriApp is preserved (apps, users,
connections, settings). The EveriApp Windows service restarts on the new
version automatically. Server functions and the Python Packages page are new
capabilities available immediately after upgrading.
Install the Builder
- Download EveriApp-Setup-win-x64-0.17.4.exe and run it (requires admin
rights — it installs a Windows service).- Windows SmartScreen may warn that the publisher is unrecognized — this
release isn't code-signed yet. Click More info → Run anyway, and verify
the download with the checksum below if you'd like.
- Windows SmartScreen may warn that the publisher is unrecognized — this
- Choose a port when prompted (default 8800).
- The installer starts the EveriApp service and opens your browser.
- Create (or sign in to) your administrator account.
- Open AI Providers, add a provider + API key, and start building.
Installs to C:\Program Files\EveriApp, runs as the EveriApp Windows
service, and stores data under C:\ProgramData\EveriApp.
Install the Agent (optional — only for remote deployments)
On each server you deploy to, run EveriApp-Agent-Setup-win-x64-0.1.1.exe,
paste the agent token (create one in EveriApp under Secrets, category
agent_token), then add a Deployment Target in EveriApp pointing at that
server's IP on port 8765.
System requirements
- Windows 10 or 11 (64-bit)
- ~2.5 GB free disk space (the bundled Python runtime + libraries add to the
download size) - 4 GB RAM minimum, 8 GB recommended
- An API key from an LLM provider (e.g. Anthropic or OpenAI) to use the builder
Verify your download (optional)
In PowerShell, in the folder where you downloaded the file:
(Get-FileHash .\EveriApp-Setup-win-x64-0.17.4.exe -Algorithm SHA256).HashIt should match the value in EveriApp-Setup-win-x64-0.17.4.exe.sha256:
6c59a41ecdfcefce8c1cc040fb4adf36c1e4dd42235db46e5ff37a77614f97db EveriApp-Setup-win-x64-0.17.4.exe
aed9063fbf4854ddec8c4946e72f438e0a9f5e3032bb2f47d4a352d5e1612c4a EveriApp-Agent-Setup-win-x64-0.1.1.exe
Known limitations
- Windows only for now — macOS and Linux desktop builds are planned.
- Not yet code-signed, so Windows SmartScreen shows an "unrecognized
publisher" prompt on first run. Verify with the SHA-256 checksum above. - Server functions run Python synchronously, one request at a time per call,
and return JSON up to 5 MiB — aggregate large results before returning them.
© EveriAI LLC