Spaces:

webmazta
/

contentai

Running

App Files Files Community

contentai / app.py

Commit History

Jul 25

00:41
Move dashboard sign-in off reserved login path

webmazta

f393d93

Jul 24

23:59
Fix Hostinger pull relay heartbeat

webmazta

74a35c2
20:58
fix: add automatic direct fallback and fast offline check for Hostinger pull relay

webmazta

4a78eed

Jul 23

02:05
Cancel pending jobs without deleting them

webmazta

f88f9b9

Jul 22

01:50
Pull WordPress uploads from Hostinger

webmazta

8f8f486

Jul 20

23:40
Remove temporary WordPress upload diagnostic

webmazta

8dc625b
23:27
Add authenticated WordPress upload diagnostic

webmazta

0b0732c
22:49
Secure dashboard authentication with Space secrets

webmazta

c05bc61

Jul 17

23:50
Add Docker app setup

webmazta

cecad47

May 6

01:05
Add Facebook Page Story publishing

webmazta

e62f96a
00:52
Throttle Facebook social publishing

webmazta

1cfda26
00:45
Retry Facebook first comments separately

webmazta

ea295c1
00:36
Harden Facebook publish timeout handling

webmazta

b10cd2a

May 5

23:53
Add Facebook social publishing queue

webmazta

50fc739

May 3

16:38
feat: add Nemotron backup model + model switcher in settings

webmazta

ba31bb0

Apr 20

01:05
Fix UI caching bug preventing javascript redo logic from executing

webmazta

324a56b

Apr 19

23:26
Enhance Redo functionality with randomized frame selection

webmazta

f829051

Apr 15

09:14
fix: stuck job queue - 10min timeout, 90s AI timeout, watchdog, force-kill button, single worker thread⁠

webmazta

c845149

Apr 11

22:18
fix: ensure output directory exists before saving featured images

webmazta

00785d3
17:53
perf: batch settings save into single transaction to prevent momentary freezes on HF Spaces

webmazta

eb2a005
17:43
fix: use /data persistent volume on HF Spaces to prevent restart loops when saving settings

webmazta

7fc7746
15:55
fix: use absolute paths for DB_PATH and UPLOAD_DIR to prevent 'unable to open database file' errors on Windows

webmazta

2c1130d

Apr 9

21:46
feat: replace OpenRouter/Groq with Gemma 4 as single AI engine

webmazta

b21e6da

Apr 8

09:28
fix: recover stuck 'running' jobs on startup and /api/jobs/stop

webmazta

0c23ff5
09:25
perf: slash cooldowns to absolute minimum

webmazta

80adb7f
01:00
perf: aggressive cooldown reduction with OpenRouter credits

webmazta

23e071a
00:25
perf: reduce all cooldowns by 70% for Gemma models

webmazta

939a9ef

Apr 7

00:16
fix(app): protect deployment secrets from being overwritten by DB values

webmazta

71a3777

Apr 6

22:58
fix(app): don't let empty DB settings overwrite real env vars

webmazta

fd63d78
19:16
feat(optimizations): reduce frames, add proactive spacing, caching, and project guide

webmazta

636ec54
02:40
feat(llm): add 90s cooldown + auto-retry on OpenRouter rate limit (429)

webmazta

4b270d1
02:03
fix(jobs): add 1.5min cooldown and LOCK to _process_all_pending for strict sequential processing

webmazta

5aa3c9e
01:22
fix(deps): use venv python in startup script to resolve openai version mismatch

webmazta

12b4f45

Apr 5

21:00
fix(wp): handle bot verification blocks and fix SQLite database locking

webmazta

a8496a0
12:15
fix(ai): resolve silent AI skips by dynamically reading API keys in worker thread context

webmazta

bf884f0
11:21
Add Redo button and UI improvements (manual tab switching and initiation toasts)

webmazta

cd67bfc

Apr 4

03:56
fix: process jobs sequentially instead of 5 at a time

webmazta

56f95f3
03:25
fix: add /api/upload endpoint, groupMedia checkbox, and all missing API routes for grouped media feature

webmazta

4b4d6a5
02:10
feat: implement Grouped Media feature and refactor app.py

Webmazta

50aa7df
01:39
Fix security vulnerabilities, optimize performance, and improve error handling⁠

Webmazta

5554638

Apr 3

20:19
Complete migration of RSS screening logic to OpenRouter/Groq dual-engine

Webmazta

e49345b
12:35
Fix Gemini model names and add Dockerfile for Render deployment

Webmazta

85caec3

Mar 14

04:55
Update

Your Name

ce852fb

Mar 12

22:27
Standardized Gemini SDK to legacy google-generativeai and fixed gemini-2.5-flash compatibility

Webmazta

b23cb24
21:03
feat: add manual video and image upload support for AI analysis and WP publishing

Webmazta

b22c91c

Mar 11

01:07
Add manual review queue for RSS feeds and allow deleting failed jobs

Webmazta

eec5a0a
00:59
Fix RSS feeds persistence on Render by using a persistent disk

Webmazta

0daba2b

Mar 10

03:29
Implement RSS link memory to prevent duplicate processing

Webmazta

fb79856
03:19
Sync notification environment variables to DB settings on startup

Webmazta

d0c61d5
03:09
Add authentication (webmazta/2010), implement premium light mode, and improve mobile responsiveness

Webmazta

955f291

Commit History

Jul 25

Move dashboard sign-in off reserved login path

Jul 24

Fix Hostinger pull relay heartbeat

fix: add automatic direct fallback and fast offline check for Hostinger pull relay

Jul 23

Cancel pending jobs without deleting them

Jul 22

Pull WordPress uploads from Hostinger

Jul 20

Remove temporary WordPress upload diagnostic

Add authenticated WordPress upload diagnostic

Secure dashboard authentication with Space secrets

Jul 17

Add Docker app setup

May 6

Add Facebook Page Story publishing

Throttle Facebook social publishing

Retry Facebook first comments separately

Harden Facebook publish timeout handling

May 5

Add Facebook social publishing queue

May 3

feat: add Nemotron backup model + model switcher in settings

Apr 20

Fix UI caching bug preventing javascript redo logic from executing

Apr 19

Enhance Redo functionality with randomized frame selection

Apr 15

fix: stuck job queue - 10min timeout, 90s AI timeout, watchdog, force-kill button, single worker thread⁠

Apr 11

fix: ensure output directory exists before saving featured images

perf: batch settings save into single transaction to prevent momentary freezes on HF Spaces

fix: use /data persistent volume on HF Spaces to prevent restart loops when saving settings

fix: use absolute paths for DB_PATH and UPLOAD_DIR to prevent 'unable to open database file' errors on Windows

Apr 9

feat: replace OpenRouter/Groq with Gemma 4 as single AI engine

Apr 8

fix: recover stuck 'running' jobs on startup and /api/jobs/stop

perf: slash cooldowns to absolute minimum

perf: aggressive cooldown reduction with OpenRouter credits

perf: reduce all cooldowns by 70% for Gemma models

Apr 7

fix(app): protect deployment secrets from being overwritten by DB values

Apr 6

fix(app): don't let empty DB settings overwrite real env vars

feat(optimizations): reduce frames, add proactive spacing, caching, and project guide

feat(llm): add 90s cooldown + auto-retry on OpenRouter rate limit (429)

fix(jobs): add 1.5min cooldown and LOCK to _process_all_pending for strict sequential processing

fix(deps): use venv python in startup script to resolve openai version mismatch

Apr 5

fix(wp): handle bot verification blocks and fix SQLite database locking

fix(ai): resolve silent AI skips by dynamically reading API keys in worker thread context

Add Redo button and UI improvements (manual tab switching and initiation toasts)

Apr 4

fix: process jobs sequentially instead of 5 at a time

fix: add /api/upload endpoint, groupMedia checkbox, and all missing API routes for grouped media feature

feat: implement Grouped Media feature and refactor app.py

Fix security vulnerabilities, optimize performance, and improve error handling⁠

Apr 3

Complete migration of RSS screening logic to OpenRouter/Groq dual-engine

Fix Gemini model names and add Dockerfile for Render deployment

Mar 14

Update

Mar 12

Standardized Gemini SDK to legacy google-generativeai and fixed gemini-2.5-flash compatibility

feat: add manual video and image upload support for AI analysis and WP publishing

Mar 11

Add manual review queue for RSS feeds and allow deleting failed jobs

Fix RSS feeds persistence on Render by using a persistent disk

Mar 10

Implement RSS link memory to prevent duplicate processing

Sync notification environment variables to DB settings on startup

Add authentication (webmazta/2010), implement premium light mode, and improve mobile responsiveness