Simplify Replit deployment

This commit is contained in:
hwhung
2026-05-06 00:04:11 +08:00
parent 5463862d32
commit da856261f3
5 changed files with 112 additions and 20 deletions
+6
View File
@@ -0,0 +1,6 @@
run = "pnpm dev"
onBoot = "pnpm install --ignore-workspace"
[deployment]
build = "pnpm install --ignore-workspace && pnpm run build"
run = "pnpm start"
+28 -10
View File
@@ -4,6 +4,20 @@ This app is already prepared for one-shot deployment. Follow these commands exac
Do not add this repo to a parent workspace. Do not move it into `artifacts/`. Do not edit source files. Do not edit `package.json`. Do not create `config.json` or a `/tmp` config file for deployment. Do not inspect or free unrelated ports. Do not add this repo to a parent workspace. Do not move it into `artifacts/`. Do not edit source files. Do not edit `package.json`. Do not create `config.json` or a `/tmp` config file for deployment. Do not inspect or free unrelated ports.
## Replit
The repo includes `.replit`, so after import the Run button starts the app with:
```bash
pnpm dev
```
The `.replit` `onBoot` command installs dependencies with:
```bash
pnpm install --ignore-workspace
```
## Install ## Install
Run from this repository root: Run from this repository root:
@@ -16,37 +30,41 @@ Use the available `pnpm`. This repo intentionally has no `packageManager` pin.
## Start ## Start
Without Replit AI Integrations:
```bash ```bash
pnpm dev pnpm dev
``` ```
With Replit AI Integrations: When Replit AI Integrations are provisioned, the app reads these variables automatically:
- `AI_INTEGRATIONS_OPENAI_BASE_URL`
- `AI_INTEGRATIONS_OPENAI_API_KEY`
- `AI_INTEGRATIONS_ANTHROPIC_BASE_URL`
- `AI_INTEGRATIONS_ANTHROPIC_API_KEY`
Provider CLI flags are still supported for manual overrides:
```bash ```bash
pnpm dev -- \ pnpm dev -- \
--host 0.0.0.0 \
--port 3000 \
--proxy-api-key dev-proxy-key \
--openai-base-url "<ai-integrations-openai-base-url>" \ --openai-base-url "<ai-integrations-openai-base-url>" \
--openai-api-key "<ai-integrations-openai-api-key>" \ --openai-api-key "<ai-integrations-openai-api-key>" \
--anthropic-base-url "<ai-integrations-anthropic-base-url>" \ --anthropic-base-url "<ai-integrations-anthropic-base-url>" \
--anthropic-api-key "<ai-integrations-anthropic-api-key>" --anthropic-api-key "<ai-integrations-anthropic-api-key>"
``` ```
The provider flags above are implemented. Do not verify them by reading source first.
Use Replit AI Integrations endpoints only. Never use public official provider endpoints. Do not print provider tokens. Use Replit AI Integrations endpoints only. Never use public official provider endpoints. Do not print provider tokens.
If the process does not stay alive after the shell exits, create the smallest background workflow that runs the same `pnpm dev ...` command from this repository root.
## Defaults ## Defaults
- host: `0.0.0.0` - host: `0.0.0.0`
- port: `3000` - port: `3000`
- bearer token: `dev-proxy-key` - bearer token: `dev-proxy-key`
These can be overridden with CLI flags or environment variables:
- `--host` or `HOST`
- `--port` or `PORT`
- `--proxy-api-key` or `PROXY_API_KEY`
## Verify ## Verify
```bash ```bash
+22 -2
View File
@@ -2,7 +2,7 @@
## Overview ## Overview
Standalone TypeScript API server with CLI-only deployment configuration. Standalone TypeScript API server with Replit run and deployment configuration.
## Stack ## Stack
@@ -14,7 +14,14 @@ Standalone TypeScript API server with CLI-only deployment configuration.
## Provider Endpoints ## Provider Endpoints
Use Replit AI Integrations endpoints through CLI flags: Use Replit AI Integrations endpoints. The app reads these environment variables automatically:
- `AI_INTEGRATIONS_OPENAI_BASE_URL`
- `AI_INTEGRATIONS_OPENAI_API_KEY`
- `AI_INTEGRATIONS_ANTHROPIC_BASE_URL`
- `AI_INTEGRATIONS_ANTHROPIC_API_KEY`
CLI flags are still available for manual overrides:
- `--openai-base-url` - `--openai-base-url`
- `--openai-api-key` - `--openai-api-key`
@@ -31,6 +38,19 @@ Never use public official provider endpoints.
- `pnpm run build` — typecheck and build `dist/index.mjs` - `pnpm run build` — typecheck and build `dist/index.mjs`
- `pnpm start` — run the built server - `pnpm start` — run the built server
## Replit Configuration
- `.replit` `run` uses `pnpm dev`
- `.replit` `onBoot` uses `pnpm install --ignore-workspace`
- deployment build uses `pnpm install --ignore-workspace && pnpm run build`
- deployment run uses `pnpm start`
## Server Overrides
- `--host` or `HOST`
- `--port` or `PORT`
- `--proxy-api-key` or `PROXY_API_KEY`
## Routes ## Routes
- `/` and `/dashboard` — static model dashboard - `/` and `/dashboard` — static model dashboard
+34
View File
@@ -0,0 +1,34 @@
Clone and run https://git.komnenos.me/springfw/random-proxy in ./random-proxy.
Goal: get the app running once, with Replit AI Integrations wired, on port 3000.
Do not:
- add it to a parent workspace
- move it into artifacts/
- modify source files
- edit package.json
- create config.json or /tmp config files
- inspect/free unrelated ports
- use nohup/setsid/background shell tricks
- run chat-completion smoke tests unless asked
Steps:
1. Clone into ./random-proxy.
2. Provision/use Replit AI Integrations for:
- OpenAI
- Anthropic
3. Open ./random-proxy and use the configured Replit Run button.
4. Wait briefly, then verify:
curl -s http://localhost:3000/api/healthz
curl -s http://localhost:3000/dashboard | head
curl -s -H "Authorization: Bearer dev-proxy-key" http://localhost:3000/v1/models
Report only:
- local URL
- whether OpenAI and Anthropic AI Integrations were wired
- verification results
+22 -8
View File
@@ -55,31 +55,45 @@ function argFlag(name: string): boolean {
return process.argv.includes(`--${name}`); return process.argv.includes(`--${name}`);
} }
function envValue(name: string): string | undefined {
const value = process.env[name];
return value && value.length > 0 ? value : undefined;
}
function applyArgOverrides(base: AppConfig): AppConfig { function applyArgOverrides(base: AppConfig): AppConfig {
const rawPort = argValue("port"); const rawPort = argValue("port") ?? envValue("PORT");
const port = rawPort === undefined ? base.server.port : Number(rawPort); const port = rawPort === undefined ? base.server.port : Number(rawPort);
if (!Number.isInteger(port) || port <= 0) { if (!Number.isInteger(port) || port <= 0) {
throw new Error(`Invalid CLI value "--port": ${rawPort}`); throw new Error(`Invalid port value: ${rawPort}`);
} }
const host = argValue("host") ?? envValue("HOST") ?? base.server.host;
const proxyApiKey = argValue("proxy-api-key") ?? envValue("PROXY_API_KEY") ?? base.proxy.apiKey;
const openaiBaseURL = argValue("openai-base-url") ?? envValue("AI_INTEGRATIONS_OPENAI_BASE_URL");
const openaiApiKey = argValue("openai-api-key") ?? envValue("AI_INTEGRATIONS_OPENAI_API_KEY");
const anthropicBaseURL =
argValue("anthropic-base-url") ?? envValue("AI_INTEGRATIONS_ANTHROPIC_BASE_URL");
const anthropicApiKey =
argValue("anthropic-api-key") ?? envValue("AI_INTEGRATIONS_ANTHROPIC_API_KEY");
return { return {
server: { server: {
host: argValue("host") ?? base.server.host, host,
port, port,
}, },
proxy: { proxy: {
apiKey: argValue("proxy-api-key") ?? base.proxy.apiKey, apiKey: proxyApiKey,
}, },
providers: { providers: {
openai: { openai: {
...base.providers.openai, ...base.providers.openai,
...(argValue("openai-base-url") ? { baseURL: argValue("openai-base-url") } : {}), ...(openaiBaseURL ? { baseURL: openaiBaseURL } : {}),
...(argValue("openai-api-key") ? { apiKey: argValue("openai-api-key") } : {}), ...(openaiApiKey ? { apiKey: openaiApiKey } : {}),
}, },
anthropic: { anthropic: {
...base.providers.anthropic, ...base.providers.anthropic,
...(argValue("anthropic-base-url") ? { baseURL: argValue("anthropic-base-url") } : {}), ...(anthropicBaseURL ? { baseURL: anthropicBaseURL } : {}),
...(argValue("anthropic-api-key") ? { apiKey: argValue("anthropic-api-key") } : {}), ...(anthropicApiKey ? { apiKey: anthropicApiKey } : {}),
}, },
}, },
logging: { logging: {