loopsmith-cli 0.2.2__tar.gz → 0.3.1__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: loopsmith-cli
3
- Version: 0.2.2
3
+ Version: 0.3.1
4
4
  Summary: Self-evolving agent loops behind a deterministic verification gate. Installs the prebuilt Rust binary as `loopsmith`.
5
5
  Author-email: Vaibhav Agarwal <bitphill@duck.com>
6
6
  License-Expression: MIT
@@ -21,7 +21,7 @@ Requires-Python: >=3.8
21
21
  Description-Content-Type: text/markdown
22
22
 
23
23
  <div align="center">
24
- <img src="https://raw.githubusercontent.com/bitphill/loopsmith/v0.2.2/assets/loopsmith-logo-256.png" alt="loopsmith" width="180" />
24
+ <img src="https://raw.githubusercontent.com/bitphill/loopsmith/v0.3.1/assets/loopsmith-logo-256.png" alt="loopsmith" width="180" />
25
25
  <h1>loopsmith</h1>
26
26
  <p><em>Self-evolving agent loops. The gate is code, so "done" cannot be argued.</em></p>
27
27
  </div>
@@ -54,15 +54,29 @@ a goal that stops being true stops being satisfied.
54
54
 
55
55
  **Not a developer?** Marketing, sales, research, ops — if you can edit a text
56
56
  file, you can run a loop. And if you would rather not open a text file at all,
57
- `loopsmith --web` builds one for you in a browser.
57
+ two guided front ends build one *for* you by asking questions, one at a time,
58
+ with every answer explained as you go:
58
59
 
59
- ### ➜ [START-HERE — README-FOR-DUMMIES.md](https://github.com/bitphill/loopsmith/blob/v0.2.2/README-FOR-DUMMIES.md)
60
+ - **`loopsmith --web`** — in your browser. It asks whether you are new to this
61
+ first; say yes and it walks you through the whole thing, starting from a
62
+ ready-made loop if you want one.
63
+ - **`loopsmith --guided`** — the same questions in the terminal, for a machine
64
+ with no browser.
65
+
66
+ Neither one runs anything or spends anything while you answer. You are writing a
67
+ description; nothing happens until you press a button.
68
+
69
+ ### ➜ [START-HERE — README-FOR-DUMMIES.md](https://github.com/bitphill/loopsmith/blob/v0.3.1/README-FOR-DUMMIES.md)
60
70
 
61
71
  A plain-English guide: one install line, thirteen ready-made loops to copy, the
62
72
  six settings you actually edit, and how to leave it running on a schedule.
63
73
 
64
- There is also a generated [code wiki](https://bitphill.github.io/loopsmith/wiki/#overview) mapping the crates,
65
- the execution engine, the gate, and the provider layer.
74
+ There is also a generated **code wiki** mapping the crates, the execution engine,
75
+ the gate, and the provider layer — the same pages on two surfaces, whichever you
76
+ prefer to read in:
77
+
78
+ - [Browsable viewer](https://bitphill.github.io/loopsmith/wiki/#overview) — searchable, with the module map
79
+ - [GitHub Wiki](https://github.com/bitphill/loopsmith/wiki) — the same pages in the repository's Wiki tab
66
80
 
67
81
  ## What it is
68
82
 
@@ -79,6 +93,26 @@ One rule holds the whole design up:
79
93
  the gate can **revoke**: delete a required artifact and a satisfied goal flips
80
94
  back. A system that can only promote is a burndown chart with extra steps.
81
95
 
96
+ ## Guided setup, in the terminal
97
+
98
+ ```bash
99
+ loopsmith --guided # or: loopsmith guided
100
+ ```
101
+
102
+ The same config the browser paints, built by answering one question at a time —
103
+ no browser needed, so it works over SSH and in any bare terminal. Each field
104
+ explains itself and shows its default in `[brackets]`, so pressing Enter all the
105
+ way through gives you something that works. Choices are numbered: you type the
106
+ number, or the name. The agent CLIs already installed on your machine are offered
107
+ as a menu, pre-filled, so a provider is one keystroke rather than a remembered
108
+ command line.
109
+
110
+ Four commands work at every prompt: `:back` to change the last answer, `:next` to
111
+ keep the default, `:help` to explain the field again, and `:quit` to leave with a
112
+ draft you can resume. Nothing is written until the finished config passes the same
113
+ check `loopsmith validate` runs on a hand-written file, and
114
+ `loopsmith guided --edit <file>` walks through changing a loop you already have.
115
+
82
116
  ## The browser UI
83
117
 
84
118
  ```bash
@@ -93,6 +127,29 @@ Six steps rather than one long form — Place, Power, Intent, Proof, Work, Ship
93
127
  carrying only the actions that make sense on each, and `⌘K` to reach any step,
94
128
  section, action, or example directly.
95
129
 
130
+ ### Two ways in
131
+
132
+ The first thing `--web` asks is which kind of smith you are, and it remembers the
133
+ answer:
134
+
135
+ - **an experienced smith** goes straight to the six-step editor above;
136
+ - **a new smith** gets the explanation, then the working examples to start from —
137
+ or an empty config — and then the **same one-question-at-a-time walk-through
138
+ `--guided` runs in the terminal**, drawn as cards.
139
+
140
+ The walk-through asks the sections in the order `--guided` asks them: identity,
141
+ providers, goals, validations, stop gates, then each advanced section (A–J)
142
+ behind its own opt-in card. One field per card, with that field's explanation in
143
+ place. A repeating section — goals, validations, nodes — collects entries in a
144
+ single card with a `+`, and stays there until you say **This part is done**. A
145
+ choice is a row of options you can see all of at once rather than a numbered
146
+ list. Loading an example fills the answers in without skipping the questions.
147
+
148
+ Both modes are two views of one draft: **Expert editor** on any card hands the
149
+ half-filled config to the six-step form, `⌘K` switches back, and the review panel
150
+ watches either way. Nothing is created until the real validator reports no
151
+ errors.
152
+
96
153
  It probes the machine first, so nothing has to be typed from memory: agent CLIs
97
154
  on `PATH`, the Ollama models actually pulled, MCP servers already configured by
98
155
  your editor, which API keys are set, and which sub-agents are installed. Found
@@ -1,5 +1,5 @@
1
1
  <div align="center">
2
- <img src="https://raw.githubusercontent.com/bitphill/loopsmith/v0.2.2/assets/loopsmith-logo-256.png" alt="loopsmith" width="180" />
2
+ <img src="https://raw.githubusercontent.com/bitphill/loopsmith/v0.3.1/assets/loopsmith-logo-256.png" alt="loopsmith" width="180" />
3
3
  <h1>loopsmith</h1>
4
4
  <p><em>Self-evolving agent loops. The gate is code, so "done" cannot be argued.</em></p>
5
5
  </div>
@@ -32,15 +32,29 @@ a goal that stops being true stops being satisfied.
32
32
 
33
33
  **Not a developer?** Marketing, sales, research, ops — if you can edit a text
34
34
  file, you can run a loop. And if you would rather not open a text file at all,
35
- `loopsmith --web` builds one for you in a browser.
35
+ two guided front ends build one *for* you by asking questions, one at a time,
36
+ with every answer explained as you go:
36
37
 
37
- ### ➜ [START-HERE — README-FOR-DUMMIES.md](https://github.com/bitphill/loopsmith/blob/v0.2.2/README-FOR-DUMMIES.md)
38
+ - **`loopsmith --web`** — in your browser. It asks whether you are new to this
39
+ first; say yes and it walks you through the whole thing, starting from a
40
+ ready-made loop if you want one.
41
+ - **`loopsmith --guided`** — the same questions in the terminal, for a machine
42
+ with no browser.
43
+
44
+ Neither one runs anything or spends anything while you answer. You are writing a
45
+ description; nothing happens until you press a button.
46
+
47
+ ### ➜ [START-HERE — README-FOR-DUMMIES.md](https://github.com/bitphill/loopsmith/blob/v0.3.1/README-FOR-DUMMIES.md)
38
48
 
39
49
  A plain-English guide: one install line, thirteen ready-made loops to copy, the
40
50
  six settings you actually edit, and how to leave it running on a schedule.
41
51
 
42
- There is also a generated [code wiki](https://bitphill.github.io/loopsmith/wiki/#overview) mapping the crates,
43
- the execution engine, the gate, and the provider layer.
52
+ There is also a generated **code wiki** mapping the crates, the execution engine,
53
+ the gate, and the provider layer — the same pages on two surfaces, whichever you
54
+ prefer to read in:
55
+
56
+ - [Browsable viewer](https://bitphill.github.io/loopsmith/wiki/#overview) — searchable, with the module map
57
+ - [GitHub Wiki](https://github.com/bitphill/loopsmith/wiki) — the same pages in the repository's Wiki tab
44
58
 
45
59
  ## What it is
46
60
 
@@ -57,6 +71,26 @@ One rule holds the whole design up:
57
71
  the gate can **revoke**: delete a required artifact and a satisfied goal flips
58
72
  back. A system that can only promote is a burndown chart with extra steps.
59
73
 
74
+ ## Guided setup, in the terminal
75
+
76
+ ```bash
77
+ loopsmith --guided # or: loopsmith guided
78
+ ```
79
+
80
+ The same config the browser paints, built by answering one question at a time —
81
+ no browser needed, so it works over SSH and in any bare terminal. Each field
82
+ explains itself and shows its default in `[brackets]`, so pressing Enter all the
83
+ way through gives you something that works. Choices are numbered: you type the
84
+ number, or the name. The agent CLIs already installed on your machine are offered
85
+ as a menu, pre-filled, so a provider is one keystroke rather than a remembered
86
+ command line.
87
+
88
+ Four commands work at every prompt: `:back` to change the last answer, `:next` to
89
+ keep the default, `:help` to explain the field again, and `:quit` to leave with a
90
+ draft you can resume. Nothing is written until the finished config passes the same
91
+ check `loopsmith validate` runs on a hand-written file, and
92
+ `loopsmith guided --edit <file>` walks through changing a loop you already have.
93
+
60
94
  ## The browser UI
61
95
 
62
96
  ```bash
@@ -71,6 +105,29 @@ Six steps rather than one long form — Place, Power, Intent, Proof, Work, Ship
71
105
  carrying only the actions that make sense on each, and `⌘K` to reach any step,
72
106
  section, action, or example directly.
73
107
 
108
+ ### Two ways in
109
+
110
+ The first thing `--web` asks is which kind of smith you are, and it remembers the
111
+ answer:
112
+
113
+ - **an experienced smith** goes straight to the six-step editor above;
114
+ - **a new smith** gets the explanation, then the working examples to start from —
115
+ or an empty config — and then the **same one-question-at-a-time walk-through
116
+ `--guided` runs in the terminal**, drawn as cards.
117
+
118
+ The walk-through asks the sections in the order `--guided` asks them: identity,
119
+ providers, goals, validations, stop gates, then each advanced section (A–J)
120
+ behind its own opt-in card. One field per card, with that field's explanation in
121
+ place. A repeating section — goals, validations, nodes — collects entries in a
122
+ single card with a `+`, and stays there until you say **This part is done**. A
123
+ choice is a row of options you can see all of at once rather than a numbered
124
+ list. Loading an example fills the answers in without skipping the questions.
125
+
126
+ Both modes are two views of one draft: **Expert editor** on any card hands the
127
+ half-filled config to the six-step form, `⌘K` switches back, and the review panel
128
+ watches either way. Nothing is created until the real validator reports no
129
+ errors.
130
+
74
131
  It probes the machine first, so nothing has to be typed from memory: agent CLIs
75
132
  on `PATH`, the Ollama models actually pulled, MCP servers already configured by
76
133
  your editor, which API keys are set, and which sub-agents are installed. Found
@@ -6,7 +6,7 @@ build-backend = "setuptools.build_meta"
6
6
  # `loopsmith` on PyPI was already registered by an unrelated project, so the
7
7
  # distribution is `loopsmith-cli`. The installed command is `loopsmith`.
8
8
  name = "loopsmith-cli"
9
- version = "0.2.2"
9
+ version = "0.3.1"
10
10
  description = "Self-evolving agent loops behind a deterministic verification gate. Installs the prebuilt Rust binary as `loopsmith`."
11
11
  readme = "README.md"
12
12
  license = "MIT"
@@ -25,7 +25,7 @@ import urllib.request
25
25
  import zipfile
26
26
  from pathlib import Path
27
27
 
28
- __version__ = "0.2.2"
28
+ __version__ = "0.3.1"
29
29
 
30
30
  REPO = "bitphill/loopsmith"
31
31
  _RELEASE_BASE = f"https://github.com/{REPO}/releases/download/v{__version__}"
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: loopsmith-cli
3
- Version: 0.2.2
3
+ Version: 0.3.1
4
4
  Summary: Self-evolving agent loops behind a deterministic verification gate. Installs the prebuilt Rust binary as `loopsmith`.
5
5
  Author-email: Vaibhav Agarwal <bitphill@duck.com>
6
6
  License-Expression: MIT
@@ -21,7 +21,7 @@ Requires-Python: >=3.8
21
21
  Description-Content-Type: text/markdown
22
22
 
23
23
  <div align="center">
24
- <img src="https://raw.githubusercontent.com/bitphill/loopsmith/v0.2.2/assets/loopsmith-logo-256.png" alt="loopsmith" width="180" />
24
+ <img src="https://raw.githubusercontent.com/bitphill/loopsmith/v0.3.1/assets/loopsmith-logo-256.png" alt="loopsmith" width="180" />
25
25
  <h1>loopsmith</h1>
26
26
  <p><em>Self-evolving agent loops. The gate is code, so "done" cannot be argued.</em></p>
27
27
  </div>
@@ -54,15 +54,29 @@ a goal that stops being true stops being satisfied.
54
54
 
55
55
  **Not a developer?** Marketing, sales, research, ops — if you can edit a text
56
56
  file, you can run a loop. And if you would rather not open a text file at all,
57
- `loopsmith --web` builds one for you in a browser.
57
+ two guided front ends build one *for* you by asking questions, one at a time,
58
+ with every answer explained as you go:
58
59
 
59
- ### ➜ [START-HERE — README-FOR-DUMMIES.md](https://github.com/bitphill/loopsmith/blob/v0.2.2/README-FOR-DUMMIES.md)
60
+ - **`loopsmith --web`** — in your browser. It asks whether you are new to this
61
+ first; say yes and it walks you through the whole thing, starting from a
62
+ ready-made loop if you want one.
63
+ - **`loopsmith --guided`** — the same questions in the terminal, for a machine
64
+ with no browser.
65
+
66
+ Neither one runs anything or spends anything while you answer. You are writing a
67
+ description; nothing happens until you press a button.
68
+
69
+ ### ➜ [START-HERE — README-FOR-DUMMIES.md](https://github.com/bitphill/loopsmith/blob/v0.3.1/README-FOR-DUMMIES.md)
60
70
 
61
71
  A plain-English guide: one install line, thirteen ready-made loops to copy, the
62
72
  six settings you actually edit, and how to leave it running on a schedule.
63
73
 
64
- There is also a generated [code wiki](https://bitphill.github.io/loopsmith/wiki/#overview) mapping the crates,
65
- the execution engine, the gate, and the provider layer.
74
+ There is also a generated **code wiki** mapping the crates, the execution engine,
75
+ the gate, and the provider layer — the same pages on two surfaces, whichever you
76
+ prefer to read in:
77
+
78
+ - [Browsable viewer](https://bitphill.github.io/loopsmith/wiki/#overview) — searchable, with the module map
79
+ - [GitHub Wiki](https://github.com/bitphill/loopsmith/wiki) — the same pages in the repository's Wiki tab
66
80
 
67
81
  ## What it is
68
82
 
@@ -79,6 +93,26 @@ One rule holds the whole design up:
79
93
  the gate can **revoke**: delete a required artifact and a satisfied goal flips
80
94
  back. A system that can only promote is a burndown chart with extra steps.
81
95
 
96
+ ## Guided setup, in the terminal
97
+
98
+ ```bash
99
+ loopsmith --guided # or: loopsmith guided
100
+ ```
101
+
102
+ The same config the browser paints, built by answering one question at a time —
103
+ no browser needed, so it works over SSH and in any bare terminal. Each field
104
+ explains itself and shows its default in `[brackets]`, so pressing Enter all the
105
+ way through gives you something that works. Choices are numbered: you type the
106
+ number, or the name. The agent CLIs already installed on your machine are offered
107
+ as a menu, pre-filled, so a provider is one keystroke rather than a remembered
108
+ command line.
109
+
110
+ Four commands work at every prompt: `:back` to change the last answer, `:next` to
111
+ keep the default, `:help` to explain the field again, and `:quit` to leave with a
112
+ draft you can resume. Nothing is written until the finished config passes the same
113
+ check `loopsmith validate` runs on a hand-written file, and
114
+ `loopsmith guided --edit <file>` walks through changing a loop you already have.
115
+
82
116
  ## The browser UI
83
117
 
84
118
  ```bash
@@ -93,6 +127,29 @@ Six steps rather than one long form — Place, Power, Intent, Proof, Work, Ship
93
127
  carrying only the actions that make sense on each, and `⌘K` to reach any step,
94
128
  section, action, or example directly.
95
129
 
130
+ ### Two ways in
131
+
132
+ The first thing `--web` asks is which kind of smith you are, and it remembers the
133
+ answer:
134
+
135
+ - **an experienced smith** goes straight to the six-step editor above;
136
+ - **a new smith** gets the explanation, then the working examples to start from —
137
+ or an empty config — and then the **same one-question-at-a-time walk-through
138
+ `--guided` runs in the terminal**, drawn as cards.
139
+
140
+ The walk-through asks the sections in the order `--guided` asks them: identity,
141
+ providers, goals, validations, stop gates, then each advanced section (A–J)
142
+ behind its own opt-in card. One field per card, with that field's explanation in
143
+ place. A repeating section — goals, validations, nodes — collects entries in a
144
+ single card with a `+`, and stays there until you say **This part is done**. A
145
+ choice is a row of options you can see all of at once rather than a numbered
146
+ list. Loading an example fills the answers in without skipping the questions.
147
+
148
+ Both modes are two views of one draft: **Expert editor** on any card hands the
149
+ half-filled config to the six-step form, `⌘K` switches back, and the review panel
150
+ watches either way. Nothing is created until the real validator reports no
151
+ errors.
152
+
96
153
  It probes the machine first, so nothing has to be typed from memory: agent CLIs
97
154
  on `PATH`, the Ollama models actually pulled, MCP servers already configured by
98
155
  your editor, which API keys are set, and which sub-agents are installed. Found
File without changes