boredom1234/AI Code Review Action
Perform code reviews and comment on diffs using TogetherAI API. Available models: # meta-llama/Llama-3.3-70B-Instruct-Turbo $0.88 # meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo-128K $0.18 # deepseek-ai/DeepSeek-V3 $1.25 # deepseek-ai/DeepSeek-R1 $3.00 / $7.00 # meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo $3.50
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
- Runtime
- Deprecated runtime
Inputs
| name | description | required | default |
|---|---|---|---|
| GITHUB_TOKEN | GitHub token to interact with the repository. | yes | — |
| TOGETHER_API_KEY | TogetherAI API key for LLM access. | yes | — |
| TOGETHER_API_MODEL | TogetherAI model name (e.g., meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo, deepseek-ai/DeepSeek-V3). | no | meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo |
| exclude | Glob patterns to exclude files from the diff analysis | no | *.md,*.txt |
| enable_quality_metrics | Enable code quality metrics analysis | no | true |
| quality_tools | Comma-separated list of quality tools to run (eslint,pylint,etc) | no | auto |
| quality_config_paths | JSON mapping of tool names to config file paths (e.g., '{"eslint":".eslintrc.custom.json","pylint":"custom_pylintrc"}') | no | {} |
| ignore_rules | JSON mapping of tool names to arrays of rules to ignore (e.g., '{"eslint":["no-console","no-unused-vars"],"pylint":["missing-docstring"]}') | no | {} |
| ignore_files | Additional glob patterns to exclude from quality analysis (comma-separated) | no | "" |
| fail_on_quality_issues | Fail the action if quality issues are found | no | false |
| max_critical_issues | Maximum number of critical issues allowed before failing the action (set to -1 to disable) | no | 0 |
| max_warning_issues | Maximum number of warning issues allowed before failing the action (set to -1 to disable) | no | -1 |
| max_suggestion_issues | Maximum number of suggestion issues allowed before failing the action (set to -1 to disable) | no | -1 |
| comment_mode | How to handle comments: 'all' (comment on all issues), 'new' (only new issues), or 'unresolved' (new and unresolved issues) | no | all |
Outputs
no outputs