agent-gitv1 0.1.2__tar.gz → 0.1.3__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: agent-gitv1
3
- Version: 0.1.2
3
+ Version: 0.1.3
4
4
  Summary: AI-powered Git CLI using MCP + OpenAI, Gemini, or Ollama for commit workflows
5
5
  Author-email: Vijay <you@example.com>
6
6
  License: MIT
@@ -25,11 +25,16 @@ Supported providers:
25
25
  ## Features
26
26
 
27
27
  - `agent config` to set provider and model
28
+ - Smart staging with auto-grouping (UI/backend/config/etc.)
28
29
  - `agent commit` to generate commit messages from git diff
29
30
  - Multiple commit suggestions (`--suggestions`)
30
- - History-aware commit message generation (uses recent commit subjects)
31
+ - Repo-aware style learning from last ~20 commit subjects
32
+ - `agent explain` for change/intent/risk summaries
31
33
  - Ollama model auto-detection on local network (`/24`, port `11434`)
34
+ - Loading/thinking spinner during long-running operations
32
35
  - `agent push` to push to remote
36
+ - Fully LLM-driven `agent push` failure diagnosis from live git evidence
37
+ - Live command output streaming during push execution
33
38
 
34
39
  ## Prerequisites
35
40
 
@@ -72,8 +77,11 @@ Shows all available commands and options.
72
77
  ### `agent config`
73
78
  Configure provider and model.
74
79
 
80
+ When using Ollama, the tool can scan your local network and detect servers
81
+ like `http://192.168.1.35:11434`.
82
+
75
83
  ### `agent commit`
76
- Stage changes, generate suggestions, choose a message, and commit.
84
+ Smart-stage changes, generate repo-aware suggestions, choose messages, and commit.
77
85
 
78
86
  Examples:
79
87
 
@@ -81,9 +89,29 @@ Examples:
81
89
  agent commit
82
90
  agent commit --repo /path/to/repo
83
91
  agent commit --suggestions 3
92
+ agent commit --no-smart-stage
84
93
  agent commit --verbose
85
94
  ```
86
95
 
96
+ When smart staging is enabled (default), the CLI detects logical change groups and asks if it should create multiple commits.
97
+ During AI generation steps, a terminal loading spinner is shown.
98
+
99
+ ### `agent explain`
100
+ Explain current changes with:
101
+ - what changed
102
+ - why it likely changed
103
+ - risk areas
104
+ - files impacted
105
+
106
+ Example:
107
+
108
+ ```bash
109
+ agent explain
110
+ agent explain --repo /path/to/repo
111
+ ```
112
+
113
+ `agent explain` also shows a loading spinner while the LLM is analyzing the diff.
114
+
87
115
  ### `agent push`
88
116
  Push commits to remote.
89
117
 
@@ -94,6 +122,15 @@ agent push
94
122
  agent push --remote origin --branch main
95
123
  ```
96
124
 
125
+ `agent push` streams git output live in the terminal while running.
126
+
127
+ If push fails, the CLI now fetches live git evidence (`fetch`, local/remote logs,
128
+ ahead/behind counts, merge-base, branch tracking) and asks the LLM to diagnose
129
+ the exact root cause with recommended fix commands.
130
+ Diagnosis evidence commands also run live, so users can see each command and output.
131
+ Diagnosis output is plain text, PowerShell-command friendly, and auto-switches
132
+ to concise mode when evidence is clear.
133
+
97
134
  ## Environment Variables
98
135
 
99
136
  - `GEMINI_API_KEY` (Gemini)
@@ -10,11 +10,16 @@ Supported providers:
10
10
  ## Features
11
11
 
12
12
  - `agent config` to set provider and model
13
+ - Smart staging with auto-grouping (UI/backend/config/etc.)
13
14
  - `agent commit` to generate commit messages from git diff
14
15
  - Multiple commit suggestions (`--suggestions`)
15
- - History-aware commit message generation (uses recent commit subjects)
16
+ - Repo-aware style learning from last ~20 commit subjects
17
+ - `agent explain` for change/intent/risk summaries
16
18
  - Ollama model auto-detection on local network (`/24`, port `11434`)
19
+ - Loading/thinking spinner during long-running operations
17
20
  - `agent push` to push to remote
21
+ - Fully LLM-driven `agent push` failure diagnosis from live git evidence
22
+ - Live command output streaming during push execution
18
23
 
19
24
  ## Prerequisites
20
25
 
@@ -57,8 +62,11 @@ Shows all available commands and options.
57
62
  ### `agent config`
58
63
  Configure provider and model.
59
64
 
65
+ When using Ollama, the tool can scan your local network and detect servers
66
+ like `http://192.168.1.35:11434`.
67
+
60
68
  ### `agent commit`
61
- Stage changes, generate suggestions, choose a message, and commit.
69
+ Smart-stage changes, generate repo-aware suggestions, choose messages, and commit.
62
70
 
63
71
  Examples:
64
72
 
@@ -66,9 +74,29 @@ Examples:
66
74
  agent commit
67
75
  agent commit --repo /path/to/repo
68
76
  agent commit --suggestions 3
77
+ agent commit --no-smart-stage
69
78
  agent commit --verbose
70
79
  ```
71
80
 
81
+ When smart staging is enabled (default), the CLI detects logical change groups and asks if it should create multiple commits.
82
+ During AI generation steps, a terminal loading spinner is shown.
83
+
84
+ ### `agent explain`
85
+ Explain current changes with:
86
+ - what changed
87
+ - why it likely changed
88
+ - risk areas
89
+ - files impacted
90
+
91
+ Example:
92
+
93
+ ```bash
94
+ agent explain
95
+ agent explain --repo /path/to/repo
96
+ ```
97
+
98
+ `agent explain` also shows a loading spinner while the LLM is analyzing the diff.
99
+
72
100
  ### `agent push`
73
101
  Push commits to remote.
74
102
 
@@ -79,6 +107,15 @@ agent push
79
107
  agent push --remote origin --branch main
80
108
  ```
81
109
 
110
+ `agent push` streams git output live in the terminal while running.
111
+
112
+ If push fails, the CLI now fetches live git evidence (`fetch`, local/remote logs,
113
+ ahead/behind counts, merge-base, branch tracking) and asks the LLM to diagnose
114
+ the exact root cause with recommended fix commands.
115
+ Diagnosis evidence commands also run live, so users can see each command and output.
116
+ Diagnosis output is plain text, PowerShell-command friendly, and auto-switches
117
+ to concise mode when evidence is clear.
118
+
82
119
  ## Environment Variables
83
120
 
84
121
  - `GEMINI_API_KEY` (Gemini)