Poetry recently released v1.8.x containing a fix for the issue we were having earlier:
https://github.com/python-poetry/poetry/issues/8548
This means unavailable optional directory dependencies no longer break the docker build.
* Fix unmasking of `azure_endpoint` in `OpenAICredentials.get_api_access_kwargs()`
* Amend `ApiManager.get_models` to use `AzureOpenAI` client when `api_type` is set to `azure`
---------
Co-authored-by: Reinier van der Leer <pwuts@agpt.co>
* By default, allow requests originating from http://localhost:{AP_SERVER_PORT} instead of all origins
* Allow configuring allowed CORS origins through `AP_SERVER_CORS_ALLOWED_ORIGINS`
* Mitigate shell injection in `MacOSTTS._speech` implementation
* Mitigate shell command control bypassing in `execute_shell` and `execute_shell_popen` commands
- Improve implementation and docstring of `validate_command` function
---------
Co-authored-by: Reinier van der Leer <pwuts@agpt.co>
We have been seeing `AttributeError: 'NoneType' object has no attribute 'split'` in Sentry.
According to https://github.com/SergeyPirogov/webdriver_manager/issues/649, this occurs
when Chrome is not installed, or when no suitable ChromeDriver version is available for
the installed version of Chrome.
Instead of the `AttributeError`, we can print a better message for the agent and user to help them fix the issue.
---
Co-authored-by: kcze <kpczerwinski@gmail.com>
OpenAI requires the `id` property on tool calls in assistant messages.
We weren't storing it in the `AssistantChatMessage` that is created from the LLM's response,
causing an error when adding those messages back to the prompt.
Fix:
* Add `id` to `AssistantToolCall` and `AssistantToolCallDict` types in model_providers/schema.py
* Amend `_tool_calls_compat_extract_calls` to generate an ID for extracted tool calls
---
Co-authored-by: kcze <kpczerwinski@gmail.com>
OpenAI likes neither `tool_calls=[]` nor `tool_calls=None`. If no `tool_calls` are in the message, the key must be omitted.
This partially reverts commit 67bafa6302.
---
Co-authored-by: kcze <kpczerwinski@gmail.com>
In a `for item in list` loop, removing items from the list while iterating causes it to skip over the next item. Fix: refactor `interactively_revise_ai_settings` routine to use while loop for iterating over constraints, resources, and best practices.
---------
Co-authored-by: Kripanshu Jindal <polaris@Polaris.local>
Co-authored-by: Reinier van der Leer <pwuts@agpt.co>
- Added `task_cumulative_cost` and `task_total_cost` attributes to the `Step.additional_output` in the `AgentProtocolServer.execute_step` endpoint.
- Updated `agbenchmark` dependency in Agent and Forge
- Added a helper function `.app.utils.vcs_state_diverges_from_master()`. This function determines whether the relevant part of the codebase diverges from our `master`.
- Updated `.app.telemetry._setup_sentry()` to determine the default environment name using `vcs_state_diverges_from_master`.
OpenAI's newest models return JSON with markdown fences around it, breaking the `json.loads` parser.
This commit adds an `extract_list_from_response` function to json_utils/utilities.py and uses this function to replace `json.loads` in `_process_text`.
* Add Sentry integration for telemetry
- Add `sentry_sdk` dependency
- Add setup logic and config flow using `TELEMETRY_OPT_IN` environment variable
- Add app/telemetry.py with `setup_telemetry` helper routine
- Call `setup_telemetry` in `cli()` in app/cli.py
- Add `TELEMETRY_OPT_IN` to .env.template
- Add helper function `env_file_exists` and routine `set_env_config_value` to app/utils.py
- Add unit tests for `set_env_config_value` in test_utils.py
- Add prompt to startup to ask whether the user wants to enable telemetry if the env variable isn't set
* Add `capture_exception` statements for LLM parsing errors and command failures
- Change default `SMART_LLM` from `gpt-4` to `gpt-4-turbo-preview`
- Change default `FAST_LLM` from `gpt-3.5-turbo-16k` to `gpt-3.5-turbo-0125`
- Change default `EMBEDDING_MODEL` from `text-embedding-ada-002` to `text-embedding-3-small`
- Update .env.template, azure.yaml.template, and documentation accordingly
- Add `text-embedding-3-small` and `text-embedding-3-large` as `EMBEDDING_v3_S` and `EMBEDDING_v3_L` respectively
- Add `gpt-3.5-turbo-0125` as `GPT3_v4`
- Add `gpt-4-1106-vision-preview` as `GPT4_v3_VISION`
- Add GPT-4V models to info map
- Change chat model info mapping to derive info for aliases (e.g. `gpt-3.5-turbo`) from specific versions instead of the other way around
* Add `_sideload_chrome_extensions` subroutine to `open_page_in_browser` in web_selenium.py
* Sideloads uBlock Origin and I Still Don't Care About Cookies, downloading them if necessary
* Add 2-second delay to `open_page_in_browser` to allow time for handling cookie walls
Commit 956cdc7 "fix(agent/json_utils): Decode as JSON rather than Python objects" broke these unit tests because they generated "JSON" by stringifying a Python object.
* Compress steps in the prompt to reduce token usage, and to increase longevity when using models with limited context windows
* Move multiple copies of step formatting code to `Episode.format` method
* Add `EpisodicActionHistory.handle_compression` method to handle compression of new steps
