Provides guidelines for Claude Code instances fixing bugs in examples,
covering testing workflow, dependency updates, prose conventions, and
code style patterns derived from the Examples Style Manifesto.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* return result in inference example
* add LLM inference samples
* fix arg type
* mostly placate mypy, silence it
* proper testing for the endpoint example
* bump timeout on mypy
* set optimum version back to 1.26.1 since 1.27 has come out and breaks dep on transformers
* update bpy for blender example
* update timeout to handle longer running examples
* allow timeout parameter to run_example::main
* WebRTC and YOLO example (#1176)
webrtc example using aiortc and yolo
* attempt to fix frontmatter
* trying a different pattern for the cmd arg with m flag
* address review comments
* fix github links and add suggestions from warnings for huggingface hub and web_url vs get_web_url
* run precommit hooks
* move test to separate file
* move test to separate file
* why did changes from modal_webrtc not commit previously?
* change testing file name to not trigger pytest directly
* change cmd in front matter to change new file name
* fix typehint for peer class variable and move so we can use ModalWebRtcPeer as type
* rename test file and ignore in pyproject pytest options
* better signaling server doct string
* update docstrings and change how we specificy the modal peer class
* revert change in workflow from local testing
* change cmd to match test file name
* debug checking provided modal peer class is subsclass of ModalWebRtcPeer
* change negotation to signaling in copy/diagrams, remove deploy:true frontmatter
* close revision of copy
* remove print statement and add information to test fail message
* close read and revision of copy
* add region note and increase ws connection timeout
* address jacks comments
* make diagrams more modal-style
* add tag for lamda-test opt-out reason
* update fastrtc example
* add video and link to demo, edit text
---------
Co-authored-by: Ben Shababo <shababo@Bens-MacBook-Air.local>
Co-authored-by: Charles Frye <charles@modal.com>
* increment line limit to 81
* increment line length to 82
* increment line length to 83
* increment line length to 84
* increment line length to 85
* increment line length to 86
* increment line length to 87
* increment line length to 88
* incomplete WIP
* move secret into step that needs it, use script instead of action
* not .py, as module
* commit a python file to test diff workflow
* remove unneeded dummy python file
* adjust event name handling
* add test event with no changed files
* add random example execution to run_example script
* add a scheduled workflow to force build a random example
* fix cron syntax
* fix YAML format for job
* turn off automount during tests
* remove old test for text-to-pokemon naming
* turn off automounting during tests
* ignore two examples that end with _test.py
Adjustments of examples for upcoming SDK changes:
* Deprecating `Mount` as a user-facing object type
* Using `.add_local_`* methods on image to add local assets to containers
* tags up through 03
* add tags for code interpreter
* add tags for 06_gpu_and_ml
* add tags for 09_job_queues
* add tags for 10_integrations
* add tags for featured examples
* remove tag for feature-batched
* add tags to example objects
* rework vllm_inference example to showcase openai-compatible mode
* fix cmd
* remove faulty testing cmd
* remove extra copy of download
* switch to testing vllm_inference with serve
* better testing for download
* test via serving
* serve requires full path
* you dropped this: .py
* make pytest skippable via new metadata field
* skip in pytest (does a Volume lookup)
* inital draft of action to test monitoring
* cleans up monitoring workflow
* more complete draft of monitoring test action
* removes draft monitoring workflow, reorganizes existing workflows
* update internal development requirements
* turn off dry run now that we're going back to prod
* reorganize environment setup
* WIP version of example execution
* adds .secrets file from act
* handles modal serve, proper system exit, drops extra script
* updates actions, better environment setup
* handle PRs with no changed files
* add back dev dependencies for jupytext and pydantic in deploy
* reverts changes to typechecking to avoid slowdown
* remove diffusers training script example
* remove duckdb taxi and stable diffusion slackbot
* removes old alpaca example
* removes old gpt2 example
* moves batch_inference and falcon examples to misc and clarifies QoS
* removes misc from monitoring
* move requirements for stable lm into file, move into misc
* ignores files in misc
* generalize language in the s3 bucket example
* removes reference to the stable diffusion slackbot from the stable diffusion cli example
* moves webscraper example out of misc, it's useful
* updates tensorflow tutorial, pins version, drops conda
* updates ruff version to match CI
* line up CI and pre-commit for formatting
* applies latest version of ruff formatting
