ai-cr 3.1.2__tar.gz → 3.2.0__tar.gz

{ai_cr-3.1.2 → ai_cr-3.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: ai-cr
-Version: 3.1.2
+Version: 3.2.0
 Summary: AI code review tool that works with any language model provider. It detects issues in GitHub pull requests or local changes—instantly, reliably, and without vendor lock-in.
 License: MIT
 Keywords: static code analysis,code review,code quality,ai,coding,assistant,llm,github,automation,devops,developer tools,github actions,workflows,git
@@ -16,7 +16,7 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Software Development
 Requires-Dist: GitPython (>=3.1.44,<4.0.0)
-Requires-Dist: ai-microcore (==4.2.1)
+Requires-Dist: ai-microcore (==4.3.0)
 Requires-Dist: anthropic (>=0.57.1,<0.58.0)
 Requires-Dist: ghapi (>=1.0.6,<1.1.0)
 Requires-Dist: google-generativeai (>=0.8.5,<0.9.0)
@@ -37,10 +37,22 @@ Description-Content-Type: text/markdown
 </p>
 
 **Gito** is an open-source **AI code reviewer** that works with any language model provider.
-It detects issues in GitHub pull requests or local changes—instantly, reliably, and without vendor lock-in.
+It detects issues in GitHub pull requests or local codebase changes—instantly, reliably, and without vendor lock-in.
 
 Get consistent, thorough code reviews in seconds—no waiting for human availability.
 
+## 📋 Table of Contents
+- [Why Gito?](#-why-gito)
+- [Perfect For](#-perfect-for)
+- [Quickstart](#-quickstart)
+  - [1. Review Pull Requests via GitHub Actions](#1-review-pull-requests-via-github-actions)
+  - [2. Running Code Analysis Locally](#2-running-code-analysis-locally)
+- [Configuration](#-configuration)
+- [Documentation](#-documentation)
+- [Development Setup](#-development-setup)
+- [Contributing](#-contributing)
+- [License](#-license)
+
 ## ✨ Why Gito?
 
 - [⚡] **Lightning Fast:** Get detailed code reviews in seconds, not days — powered by parallelized LLM processing  
@@ -106,7 +118,7 @@ jobs:
           code-review-report.json
 ```
 
-> ⚠️ Make sure to add `LLM_API_KEY` to your repository’s GitHub secrets.
+> ⚠️ Make sure to add `LLM_API_KEY` to your repository's GitHub secrets.
 
 💪 Done!  
 PRs to your repository will now receive AI code reviews automatically. ✨  
@@ -180,6 +192,14 @@ See default configuration [here](https://github.com/Nayjest/Gito/blob/main/gito/
 
 More details can be found in [📖 Configuration Cookbook](https://github.com/Nayjest/Gito/blob/main/documentation/config_cookbook.md)
 
+## 📚 Documentation
+
+- [Command Line Reference](https://github.com/Nayjest/Gito/blob/main/documentation/command_line_reference.md)
+- [Configuration Cookbook](https://github.com/Nayjest/Gito/blob/main/documentation/config_cookbook.md)
+- [GitHub Setup Guide](https://github.com/Nayjest/Gito/blob/main/documentation/github_setup.md)
+- [Troubleshooting](https://github.com/Nayjest/Gito/blob/main/documentation/troubleshooting.md)
+
+
 ## 💻 Development Setup
 
 Install dependencies:

{ai_cr-3.1.2 → ai_cr-3.2.0}/README.md

@@ -8,10 +8,22 @@
 </p>
 
 **Gito** is an open-source **AI code reviewer** that works with any language model provider.
-It detects issues in GitHub pull requests or local changes—instantly, reliably, and without vendor lock-in.
+It detects issues in GitHub pull requests or local codebase changes—instantly, reliably, and without vendor lock-in.
 
 Get consistent, thorough code reviews in seconds—no waiting for human availability.
 
+## 📋 Table of Contents
+- [Why Gito?](#-why-gito)
+- [Perfect For](#-perfect-for)
+- [Quickstart](#-quickstart)
+  - [1. Review Pull Requests via GitHub Actions](#1-review-pull-requests-via-github-actions)
+  - [2. Running Code Analysis Locally](#2-running-code-analysis-locally)
+- [Configuration](#-configuration)
+- [Documentation](#-documentation)
+- [Development Setup](#-development-setup)
+- [Contributing](#-contributing)
+- [License](#-license)
+
 ## ✨ Why Gito?
 
 - [⚡] **Lightning Fast:** Get detailed code reviews in seconds, not days — powered by parallelized LLM processing  
@@ -77,7 +89,7 @@ jobs:
           code-review-report.json
 ```
 
-> ⚠️ Make sure to add `LLM_API_KEY` to your repository’s GitHub secrets.
+> ⚠️ Make sure to add `LLM_API_KEY` to your repository's GitHub secrets.
 
 💪 Done!  
 PRs to your repository will now receive AI code reviews automatically. ✨  
@@ -151,6 +163,14 @@ See default configuration [here](https://github.com/Nayjest/Gito/blob/main/gito/
 
 More details can be found in [📖 Configuration Cookbook](https://github.com/Nayjest/Gito/blob/main/documentation/config_cookbook.md)
 
+## 📚 Documentation
+
+- [Command Line Reference](https://github.com/Nayjest/Gito/blob/main/documentation/command_line_reference.md)
+- [Configuration Cookbook](https://github.com/Nayjest/Gito/blob/main/documentation/config_cookbook.md)
+- [GitHub Setup Guide](https://github.com/Nayjest/Gito/blob/main/documentation/github_setup.md)
+- [Troubleshooting](https://github.com/Nayjest/Gito/blob/main/documentation/troubleshooting.md)
+
+
 ## 💻 Development Setup
 
 Install dependencies:

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/bootstrap.py

@@ -35,7 +35,7 @@ def bootstrap(verbosity: int = 1):
     log_levels_by_verbosity = {
         0: logging.CRITICAL,
         1: logging.INFO,
-        2: logging.DEBUG,
+        2: logging.INFO,
         3: logging.DEBUG,
     }
     Env.verbosity = verbosity

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/cli.py

@@ -47,14 +47,31 @@ def main():
         app()
 
 
-@app.callback(invoke_without_command=True)
+@app.callback(
+    invoke_without_command=True,
+    help="\bGito is an open-source AI code reviewer that works with any language model provider."
+         "\nIt detects issues in GitHub pull requests or local codebase changes"
+         "—instantly, reliably, and without vendor lock-in."
+)
 def cli(
     ctx: typer.Context,
-    verbose: bool = typer.Option(default=None),
     verbosity: int = typer.Option(
         None,
         '--verbosity', '-v',
-        help="Set verbosity level (0-3, default 1)"
+        show_default=False,
+        help="\b"
+             "Set verbosity level. Supported values: 0-3. Default: 1."
+             "\n [ 0 ]: no additional output, "
+             "\n [ 1 ]: normal mode, shows warnings, shortened LLM requests and logging.INFO"
+             "\n [ 2 ]: verbose mode, show full LLM requests"
+             "\n [ 3 ]: very verbose mode, also debug information"
+    ),
+    verbose: bool = typer.Option(
+        default=None,
+        help="\b"
+             "--verbose is equivalent to -v2, "
+             "\n--no-verbose is equivalent to -v0. "
+             "\n(!) Can't be used together with -v or --verbosity."
     ),
 ):
     if verbose is not None and verbosity is not None:
@@ -71,7 +88,7 @@ def cli(
 
 
 @app_no_subcommand.command(name="review", help="Perform code review")
-@app.command(name="review", help="Perform code review")
+@app.command(name="review", help="Perform a code review of the target codebase changes.")
 @app.command(name="run", hidden=True)
 def cmd_review(
     refs: str = arg_refs(),
@@ -120,7 +137,7 @@ def cmd_review(
             )
 
 
-@app.command(name="ask", help="Answer questions about codebase changes")
+@app.command(name="ask", help="Answer questions about the target codebase changes.")
 @app.command(name="answer", hidden=True)
 @app.command(name="talk", hidden=True)
 def cmd_answer(
@@ -140,6 +157,10 @@ def cmd_answer(
         default=None,
         help="GitHub Pull Request number"
     ),
+    aux_files: list[str] = typer.Option(
+        default=None,
+        help="Auxiliary files that might be helpful"
+    )
 ):
     _what, _against = args_to_target(refs, what, against)
     pr = pr or os.getenv("PR_NUMBER_FROM_WORKFLOW_DISPATCH")
@@ -157,6 +178,7 @@ def cmd_answer(
         prompt_file=prompt_file,
         use_pipeline=use_pipeline,
         pr=pr,
+        aux_files=aux_files,
     )
     if post_to == 'linear':
         logging.info("Posting answer to Linear...")
@@ -164,12 +186,12 @@ def cmd_answer(
     return out
 
 
-@app.command(help="Configure LLM for local usage interactively")
+@app.command(help="Configure LLM for local usage interactively.")
 def setup():
     mc.interactive_setup(HOME_ENV_PATH)
 
 
-@app.command(name="render")
+@app.command(name="render", help="Render and display code review report.")
 @app.command(name="report", hidden=True)
 def render(
     format: str = typer.Argument(default=Report.Format.CLI),
@@ -183,7 +205,11 @@ def render(
     Report.load(file_name=source).to_cli(report_format=format)
 
 
-@app.command(help="List files in the diff. Might be useful to check what will be reviewed.")
+@app.command(
+    help="\bList files in the changeset. "
+         "\nMight be useful to check what will be reviewed if run `gito review` "
+         "with current CLI arguments and options."
+)
 def files(
     refs: str = arg_refs(),
     what: str = arg_what(),

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/cli_base.py

@@ -30,7 +30,11 @@ def args_to_target(refs, what, against) -> tuple[str | None, str | None]:
 def arg_refs() -> typer.Argument:
     return typer.Argument(
         default=None,
-        help="Git refs to review, [what]..[against] e.g. 'HEAD..HEAD~1'"
+        help=(
+            "Git refs to review, [what]..[against] (e.g., 'HEAD..HEAD~1'). "
+            "If omitted, the current index (including added but not committed files) "
+            "will be compared to the repository’s main branch."
+        ),
     )
 
 

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/commands/deploy.py

@@ -12,7 +12,10 @@ from ..cli_base import app
 from ..gh_api import gh_api
 
 
-@app.command(name="deploy", help="Deploy Gito workflows to GitHub Actions")
+@app.command(
+    name="deploy",
+    help="\bCreate and configure Gito GitHub Actions for current repository.\naliases: init"
+)
 @app.command(name="init", hidden=True)
 def deploy(
     api_type: ApiType = None,

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/commands/fix.py

@@ -15,7 +15,10 @@ from ..constants import JSON_REPORT_FILE_NAME
 from ..report_struct import Report, Issue
 
 
-@app.command(help="Fix an issue from the code review report")
+@app.command(
+    help="Fix an issue from the code review report "
+         "(latest code review results will be used by default)"
+)
 def fix(
     issue_number: int = typer.Argument(..., help="Issue number to fix"),
     report_path: Optional[str] = typer.Option(

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/commands/linear_comment.py

@@ -29,7 +29,7 @@ def post_linear_comment(issue_key, text, api_key):
     return response.json()
 
 
-@app.command()
+@app.command(help="Post a comment with specified text to the associated Linear issue.")
 def linear_comment(
     text: str = typer.Argument(None),
     refs: str = arg_refs(),

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/commands/repl.py

@@ -23,6 +23,8 @@ from ..utils import *
 from ..gh_api import *
 
 
-@app.command(help="python REPL")
+@app.command(
+    help="Python REPL with core functionality loaded for quick testing/debugging and exploration."
+)
 def repl():
     code.interact(local=globals())

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/commands/version.py

@@ -2,7 +2,7 @@ from ..cli_base import app
 from ..env import Env
 
 
-@app.command(name='version', help='Show the version of gito.bot')
+@app.command(name='version', help='Show Gito version.')
 def version():
     print(Env.gito_version)
     return Env.gito_version

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/config.toml

@@ -242,32 +242,8 @@ If there are concerns about how thoroughly the code covers the requirements and
 {{ summary_requirements -}}
 """
 answer_github_comments = true
-answer_prompt = """
-{{ self_id }}
-----TASK----
-Answer the following user question:
---USER QUESTION--
-{{ question }}
-----
-----RELATED CODEBASE CHANGES----
-{% for part in diff %}{{ part }}\n{% endfor %}
-
-----FULL FILE CONTENT AFTER APPLYING CHANGES----
-{% for file, file_lines in all_file_lines.items() %}
---FILE: {{ file }}--
-{{ file_lines }}
-{% endfor %}
-
-{%- if pipeline_out.associated_issue and pipeline_out.associated_issue.title %}
-----ASSOCIATED ISSUE----
-# {{ pipeline_out.associated_issue.title }}
-{{ pipeline_out.associated_issue.description }}
-URL: {{ pipeline_out.associated_issue.url }}
-{%- endif -%}{{ '\n' }}
-
-----ANSWERING INSTRUCTIONS----
-{{ answering_instructions }}
-"""
+answer_prompt = "tpl:answer.j2"
+aux_files = []
 [pipeline_steps.jira]
 call="gito.pipeline_steps.jira.fetch_associated_issue"
 envs=["local","gh-action"]

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/core.py

@@ -190,7 +190,7 @@ def get_diff(
                 else:
                     logging.error(f"Can't find other merge parent for {merge_sha}")
             else:
-                logging.error(
+                logging.warning(
                     f"No merge‐commit found for {current_ref!r}→{against!r}; "
                     "falling back to merge‐base diff"
                 )
@@ -243,18 +243,20 @@ def filter_diff(
     return files
 
 
-def file_lines(repo: Repo, file: str, max_tokens: int = None, use_local_files: bool = False) -> str:
+def read_file(repo: Repo, file: str, use_local_files: bool = False) -> str:
     if use_local_files:
         file_path = Path(repo.working_tree_dir) / file
         try:
-            text = file_path.read_text(encoding='utf-8')
+            return file_path.read_text(encoding='utf-8')
         except (FileNotFoundError, UnicodeDecodeError) as e:
             logging.warning(f"Could not read file {file} from working directory: {e}")
-            text = repo.tree()[file].data_stream.read().decode('utf-8')
-    else:
-        # Read from HEAD (committed version)
-        text = repo.tree()[file].data_stream.read().decode('utf-8')
 
+    # Read from HEAD (committed version)
+    return repo.tree()[file].data_stream.read().decode('utf-8')
+
+
+def file_lines(repo: Repo, file: str, max_tokens: int = None, use_local_files: bool = False) -> str:
+    text = read_file(repo=repo, file=file, use_local_files=use_local_files)
     lines = [f"{i + 1}: {line}\n" for i, line in enumerate(text.splitlines())]
     if max_tokens:
         lines, removed_qty = mc.tokenizing.fit_to_token_size(lines, max_tokens)
@@ -265,6 +267,22 @@ def file_lines(repo: Repo, file: str, max_tokens: int = None, use_local_files: b
     return "".join(lines)
 
 
+def read_files(repo: Repo, files: list[str], max_tokens: int = None) -> dict:
+    out = dict()
+    total_tokens = 0
+    for file in files:
+        content = read_file(repo=repo, file=file, use_local_files=True)
+        total_tokens += mc.tokenizing.num_tokens_from_string(file)
+        total_tokens += mc.tokenizing.num_tokens_from_string(content)
+        if max_tokens and total_tokens > max_tokens:
+            logging.warning(
+                f"Skipping file {file} due to exceeding max_tokens limit ({max_tokens})"
+            )
+            continue
+        out[file] = content
+    return out
+
+
 def make_cr_summary(ctx: Context, **kwargs) -> str:
     return (
         mc.prompt(
@@ -428,6 +446,7 @@ def answer(
     use_pipeline: bool = True,
     prompt_file: str = None,
     pr: str | int = None,
+    aux_files: list[str] = None,
 ) -> str | None:
     try:
         repo, config, diff, lines = _prepare(
@@ -454,18 +473,33 @@ def answer(
             steps=config.pipeline_steps
         )
         pipe.run()
+
+    if aux_files or config.aux_files:
+        aux_files_dict = read_files(
+            repo,
+            (aux_files or list()) + config.aux_files,
+            config.max_code_tokens // 2
+        )
+    else:
+        aux_files_dict = dict()
+
+    if not prompt_file and config.answer_prompt.startswith("tpl:"):
+        prompt_file = str(config.answer_prompt)[4:]
+
     if prompt_file:
         prompt_func = partial(mc.tpl, prompt_file)
     else:
         prompt_func = partial(mc.prompt, config.answer_prompt)
+    prompt = prompt_func(
+        question=question,
+        diff=diff,
+        all_file_lines=lines,
+        pipeline_out=ctx.pipeline_out,
+        aux_files=aux_files_dict,
+        **config.prompt_vars,
+    )
     response = mc.llm(
-        prompt_func(
-            question=question,
-            diff=diff,
-            all_file_lines=lines,
-            pipeline_out=ctx.pipeline_out,
-            **config.prompt_vars,
-        ),
+        prompt,
         callback=make_streaming_function() if Env.verbosity == 0 else None,
     )
     return response

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/project_config.py

@@ -32,6 +32,7 @@ class ProjectConfig:
     Defines the keyword or mention tag that triggers bot actions
     when referenced in code review comments.
     """
+    aux_files: list[str] = field(default_factory=list)
     pipeline_steps: dict[str, dict | PipelineStep] = field(default_factory=dict)
     collapse_previous_code_review_comments: bool = field(default=True)
     """

ai_cr-3.1.2/gito/tpl/questions/testing_guide.j2 → ai_cr-3.2.0/gito/tpl/answer.j2

@@ -1,11 +1,9 @@
 {{ self_id }}
 ----TASK----
-Create a concise comment for QA specialists that explains how to test the code-base changes.
-Include:
-- User journeys that could be disrupted by failures
-- Relevant security scenarios and edge cases
-- Potential risk areas introduced by the changes, with suggestions for focused testing to mitigate them
-
+Answer the following user question:
+--USER QUESTION--
+{{ question }}
+----
 ----RELATED CODEBASE CHANGES----
 {% for part in diff %}{{ part }}\n{% endfor %}
 
@@ -15,9 +13,14 @@ Include:
 {{ file_lines }}
 {% endfor %}
 
+{% include "partial/aux_files.j2" %}
+
 {%- if pipeline_out.associated_issue and pipeline_out.associated_issue.title %}
 ----ASSOCIATED ISSUE----
 # {{ pipeline_out.associated_issue.title }}
 {{ pipeline_out.associated_issue.description }}
 URL: {{ pipeline_out.associated_issue.url }}
 {%- endif -%}{{ '\n' }}
+
+----ANSWERING INSTRUCTIONS----
+{{ answering_instructions }}

ai_cr-3.2.0/gito/tpl/partial/aux_files.j2

@@ -0,0 +1,8 @@
+{% if aux_files %}
+----AUXILIARY INFORMATION----
+* Use if helpful for the task.
+{% for file, text in aux_files.items() %}
+--FILE: {{ file }}--
+{{ text.strip() }}
+{%- endfor -%}
+{%- endif -%}

{ai_cr-3.1.2 → ai_cr-3.2.0}/gito/tpl/questions/release_notes.j2

@@ -16,6 +16,8 @@ Avoid internal technical jargon or developer-specific details.
 {{ file_lines }}
 {% endfor %}
 
+{%- include "partial/aux_files.j2" -%}
+
 {%- if pipeline_out.associated_issue and pipeline_out.associated_issue.title %}
 ----ASSOCIATED ISSUE----
 # {{ pipeline_out.associated_issue.title }}

ai_cr-3.2.0/gito/tpl/questions/test_cases.j2

@@ -0,0 +1,37 @@
+{{ self_id }}
+----TASK----
+Create a concise list of manual QA test cases for the newly introduced codebase changes.
+The purpose of this document is to identify user journeys that could be disrupted by failures,
+cover security scenarios and edge cases, highlight potential risk areas introduced by the changes,
+and suggest focused testing to mitigate them.
+
+- Split the test cases into two categories:
+  - Critical
+  - Extended Coverage
+
+- Structure test cases using Markdown headers or bullet-point lists.
+
+- If the context includes the necessary details, describe the steps as thoroughly as possible—like for a junior QA
+  seeing this project for the first time.
+  If there are no details on how to perform a specific action, keep it abstract.
+  It is extremely important not to invent information.
+
+- Number test cases sequentially (C-01, C-02, E-01, E-02, etc.).
+
+----RELATED CODEBASE CHANGES----
+{% for part in diff %}{{ part }}\n{% endfor %}
+
+----FULL FILE CONTENT AFTER APPLYING CHANGES----
+{% for file, file_lines in all_file_lines.items() %}
+--FILE: {{ file }}--
+{{ file_lines }}
+{% endfor %}
+
+{%- include "partial/aux_files.j2" -%}
+
+{%- if pipeline_out.associated_issue and pipeline_out.associated_issue.title %}
+----ASSOCIATED ISSUE----
+# {{ pipeline_out.associated_issue.title }}
+{{ pipeline_out.associated_issue.description }}
+URL: {{ pipeline_out.associated_issue.url }}
+{%- endif -%}{{ '\n' }}

{ai_cr-3.1.2 → ai_cr-3.2.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "ai-cr"
-version = "3.1.2"
+version = "3.2.0"
 description = "AI code review tool that works with any language model provider. It detects issues in GitHub pull requests or local changes—instantly, reliably, and without vendor lock-in."
 authors = ["Nayjest <mail@vitaliy.in>"]
 readme = "README.md"
@@ -21,7 +21,7 @@ packages = [
 
 [tool.poetry.dependencies]
 python = "^3.11"
-ai-microcore = "4.2.1"
+ai-microcore = "4.3.0"
 GitPython = "^3.1.44"
 unidiff = "^0.7.5"
 google-generativeai = "^0.8.5"