Fetching metadata from the HF Docker repository... 24 days ago
-
Update shim.js
-
Update app.py
-
Update Dockerfile
-
Update app.py
-
Update shim.js
-
Update README.md
-
Update Dockerfile
-
Update requirements.txt
-
Create shim.js
-
Update app.py
-
Update app.py
-
Update app.py
-
Update Dockerfile
-
Update README.md
-
Update Dockerfile
-
Update requirements.txt
-
Update app.py
-
Update app.py
-
Rename app (1).py to app.py
-
Upload 4 files
-
Delete server.js
-
Delete package.json
-
Delete package-lock.json
-
Delete README.md
-
Delete Dockerfile
-
Delete .gitignore
-
Delete .env.example
-
Delete .dockerignore
-
Delete tools
-
Delete src
-
Delete public
-
inject current file contents + write instruction when user asks to edit
ausername-12345
-
add project file names to prompt; handle <css> and <javascript> tags
ausername-12345
-
use XML tags (<read>/<write>) instead of markdown code fences for file ops
ausername-12345
-
minimal prompt; trim history to 20 entries
ausername-12345
-
clearer system prompt with exact format examples
ausername-12345
-
strip system prompt to bare minimum for TPM limit
ausername-12345
-
match read/write inside any code block language tag
ausername-12345
-
inject project file tree into prompt; more lenient read/write regex
ausername-12345
-
switch default to llama-3.1-8b-instant
ausername-12345
-
switch to gemma2-9b-it; shorten system prompt to reduce token usage
ausername-12345
-
increase iter limit; fix prompt to push model to act; don't log intermediate file reads to UI
ausername-12345
-
replace function calling with text-parsed read/write code fence blocks
ausername-12345
-
remove function calling (unsupported by API proxy), use plain chat only
ausername-12345
-
switch default to llama-3.3-70b-versatile for better function calling support
ausername-12345
-
default to llama-3.1-8b-instant; trim system prompt and tool descriptions to reduce token overhead
ausername-12345
-
add built-in read_file/write_file/list_files tools; re-add tool_choice to API calls
ausername-12345
-
use INFERENCE_API_URL as-is, no endpoint construction
ausername-12345
-
fix endpoint construction when INFERENCE_API_URL already includes /gen/chat/completions
ausername-12345
-
add 503 fallback: retry without tools then fall back to Gen endpoint
ausername-12345