olympus-ai 4.5.13 → 4.5.14
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.
- package/.claude-plugin/plugin.json +1 -1
- package/dist/cli/index.js +63 -27
- package/dist/cli/index.js.map +1 -1
- package/dist/hooks/olympus-hooks.cjs +257 -257
- package/dist/installer/hooks.d.ts +47 -14
- package/dist/installer/hooks.d.ts.map +1 -1
- package/dist/installer/hooks.js +45 -77
- package/dist/installer/hooks.js.map +1 -1
- package/dist/installer/index.d.ts +8 -7
- package/dist/installer/index.d.ts.map +1 -1
- package/dist/installer/index.js +49 -46
- package/dist/installer/index.js.map +1 -1
- package/package.json +1 -1
- package/resources/config/risk-keywords.json +5 -5
- package/resources/rules/common/ascii-diagram-standards.md +115 -115
- package/resources/rules/common/content-validation.md +131 -131
- package/resources/rules/common/error-handling.md +430 -430
- package/resources/rules/common/markdown-formatting.md +170 -170
- package/resources/rules/common/overconfidence-prevention.md +100 -100
- package/resources/rules/common/pathway-behaviors.json +60 -60
- package/resources/rules/common/pathway-behaviors.md +100 -100
- package/resources/rules/common/process-overview.md +157 -157
- package/resources/rules/common/terminal-formatting.md +161 -161
- package/resources/rules/common/terminology.md +189 -189
- package/resources/rules/common/welcome-message.md +118 -118
- package/resources/rules/common/workflow-changes.md +285 -285
- package/resources/rules/construction/bolt-planning.md +153 -153
- package/resources/rules/construction/bolt-review.md +143 -143
- package/resources/rules/construction/build-and-test.md +527 -527
- package/resources/rules/construction/code-generation.md +414 -414
- package/resources/rules/construction/documentation.md +201 -201
- package/resources/rules/construction/functional-design.md +135 -135
- package/resources/rules/construction/infrastructure-design.md +110 -110
- package/resources/rules/construction/nfr-design.md +106 -106
- package/resources/rules/construction/nfr-requirements.md +118 -118
- package/resources/rules/construction/test-generation.md +112 -112
- package/resources/rules/core-workflow.md +196 -196
- package/resources/rules/inception/application-design.md +195 -195
- package/resources/rules/inception/bolt-planning.md +588 -588
- package/resources/rules/inception/reverse-engineering.md +354 -354
- package/resources/rules/inception/units-generation.md +505 -505
- package/resources/rules/inception/user-stories.md +527 -527
- package/resources/rules/inception/workspace-detection.md +82 -82
- package/resources/rules/operations/operations.md +19 -19
- package/resources/skills/brief/templates/ai-dlc-intent-brief-template.md +149 -149
- package/resources/skills/getting-started/SKILL.md +79 -79
- package/resources/templates/construction/bolt-spec-template.md +270 -270
- package/resources/templates/inception/unit-brief-template.md +188 -188
- package/resources/templates/inception/units-template.md +99 -99
|
@@ -1,161 +1,161 @@
|
|
|
1
|
-
# Terminal Output Formatting Standards
|
|
2
|
-
|
|
3
|
-
Claude Code renders markdown via an Ink/React TUI renderer with ANSI escape codes. Plain paragraphs are visually invisible — they blend into the response stream. These rules ensure critical information is scannable.
|
|
4
|
-
|
|
5
|
-
## The Core Problem
|
|
6
|
-
|
|
7
|
-
In a terminal, the user sees a continuous stream of white text. Only these elements create visual differentiation:
|
|
8
|
-
|
|
9
|
-
| Element | Renders As | Use For |
|
|
10
|
-
|---------|-----------|---------|
|
|
11
|
-
| `# H1` / `## H2` | Colored text (red/orange) | Section breaks, banners |
|
|
12
|
-
| `**bold**` | Bold weight | Key terms, labels |
|
|
13
|
-
| `` `inline code` `` | Distinct background/color | Commands, file paths, values |
|
|
14
|
-
| `> blockquote` | Indented with left border | Callouts, important notes |
|
|
15
|
-
| `---` | Horizontal rule | Visual fences/separators |
|
|
16
|
-
| Emojis | Colored glyphs | Status indicators, anchors |
|
|
17
|
-
| Code blocks | Syntax-highlighted box | Code, structured data |
|
|
18
|
-
|
|
19
|
-
**Rule: Anything the user must act on or notice MUST use at least two of these elements. Never deliver critical info as a plain paragraph.**
|
|
20
|
-
|
|
21
|
-
---
|
|
22
|
-
|
|
23
|
-
## Emoji Vocabulary (Standardized)
|
|
24
|
-
|
|
25
|
-
Use consistently so users learn the visual language:
|
|
26
|
-
|
|
27
|
-
| Emoji | Meaning | When to Use |
|
|
28
|
-
|-------|---------|-------------|
|
|
29
|
-
| 🚀 | Workflow start / launch | Welcome banners, phase kickoffs |
|
|
30
|
-
| 📌 | Action required | User needs to do something NOW |
|
|
31
|
-
| ✅ | Success / complete | Task done, tests pass, validation pass |
|
|
32
|
-
| ⚠️ | Warning / attention | Non-blocking but important |
|
|
33
|
-
| ❌ | Error / failure | Something broke, blocking issue |
|
|
34
|
-
| 🔍 | Context / analysis | Showing inferred or extracted info |
|
|
35
|
-
| 📋 | Summary / recap | Phase summaries, status reports |
|
|
36
|
-
| 💡 | Recommendation / tip | AI suggestions, best practices |
|
|
37
|
-
| ⏳ | In progress / waiting | Long-running operations |
|
|
38
|
-
| 🔄 | Iteration / retry | Re-running, looping back |
|
|
39
|
-
|
|
40
|
-
**Do NOT use emojis in markdown headers that will be written to files** (breaks TOCs and parsers). Emojis are for terminal display output only.
|
|
41
|
-
|
|
42
|
-
---
|
|
43
|
-
|
|
44
|
-
## Output Patterns
|
|
45
|
-
|
|
46
|
-
### Pattern 1: Welcome / Phase Banner
|
|
47
|
-
|
|
48
|
-
```markdown
|
|
49
|
-
---
|
|
50
|
-
## 🚀 Welcome to AI-DLC! Starting a new workflow for Group 1B
|
|
51
|
-
|
|
52
|
-
**Workflow:** `group-1b-test-quality` — Test Quality & Traceability
|
|
53
|
-
**Builds on:** Group 1A test infrastructure (confirmed ✅)
|
|
54
|
-
---
|
|
55
|
-
```
|
|
56
|
-
|
|
57
|
-
Why it works: `---` fences create a visual box. `##` renders in color. Emoji grabs the eye. Bold labels + inline code for scannable key-value pairs.
|
|
58
|
-
|
|
59
|
-
---
|
|
60
|
-
|
|
61
|
-
### Pattern 2: Action Required (MOST IMPORTANT)
|
|
62
|
-
|
|
63
|
-
Use this whenever the user needs to do something. This is the pattern that matters most.
|
|
64
|
-
|
|
65
|
-
```markdown
|
|
66
|
-
---
|
|
67
|
-
## 📌 Action Required
|
|
68
|
-
|
|
69
|
-
Fill in the `[Answer]:` tags in the file below, then say **"done"** to continue.
|
|
70
|
-
|
|
71
|
-
> For multi-select questions, provide comma-separated letters (e.g., `A, B, E`).
|
|
72
|
-
|
|
73
|
-
**File:** `aidlc-docs/group-1b-test-quality/inception/intent-questions.md`
|
|
74
|
-
---
|
|
75
|
-
```
|
|
76
|
-
|
|
77
|
-
Why it works: The `📌 Action Required` header is an unmissable colored+emoji signal. The command uses `inline code` and `**bold**`. The secondary instruction lives in a `> blockquote` so it's visually subordinate but still distinct. File path is in code formatting.
|
|
78
|
-
|
|
79
|
-
---
|
|
80
|
-
|
|
81
|
-
### Pattern 3: Status / Progress Update
|
|
82
|
-
|
|
83
|
-
```markdown
|
|
84
|
-
✅ All 3,360 tests pass — Group 1A prerequisite confirmed.
|
|
85
|
-
|
|
86
|
-
⏳ Generating intent questions file...
|
|
87
|
-
```
|
|
88
|
-
|
|
89
|
-
Why it works: Emoji-first lines create a scannable status log. Each line starts with a visual indicator.
|
|
90
|
-
|
|
91
|
-
---
|
|
92
|
-
|
|
93
|
-
### Pattern 4: Validation Summary
|
|
94
|
-
|
|
95
|
-
```markdown
|
|
96
|
-
## 📋 Validation Summary
|
|
97
|
-
|
|
98
|
-
| Check | Status | Detail |
|
|
99
|
-
|-------|--------|--------|
|
|
100
|
-
| Tests pass | ✅ | 3,360 / 3,360 |
|
|
101
|
-
| Prerequisites met | ✅ | Group 1A complete |
|
|
102
|
-
| Manifest valid | ⚠️ | Missing `coverage-config.json` |
|
|
103
|
-
```
|
|
104
|
-
|
|
105
|
-
Why it works: Tables with emoji status create an at-a-glance dashboard. The `##` header separates it from surrounding text.
|
|
106
|
-
|
|
107
|
-
---
|
|
108
|
-
|
|
109
|
-
### Pattern 5: Error / Blocking Issue
|
|
110
|
-
|
|
111
|
-
```markdown
|
|
112
|
-
---
|
|
113
|
-
## ❌ Blocking Issue
|
|
114
|
-
|
|
115
|
-
**What failed:** Contract validation for `UserService.create()`
|
|
116
|
-
**Why:** Missing required field `email` in test fixture at `tests/fixtures/users.ts:42`
|
|
117
|
-
|
|
118
|
-
> Fix the fixture and re-run with **"retry"**.
|
|
119
|
-
---
|
|
120
|
-
```
|
|
121
|
-
|
|
122
|
-
---
|
|
123
|
-
|
|
124
|
-
### Pattern 6: Context Extraction / Inference Review
|
|
125
|
-
|
|
126
|
-
```markdown
|
|
127
|
-
## 🔍 Context Extracted from Your Input
|
|
128
|
-
|
|
129
|
-
- **Problem:** AI-generated tests have documented failure modes...
|
|
130
|
-
- **Primary users:** Olympus OSS users running AI-DLC workflows...
|
|
131
|
-
- **Scope:** Four capabilities — anti-pattern detection, traceability...
|
|
132
|
-
- **Success criteria:** Each capability integrates into the pipeline...
|
|
133
|
-
|
|
134
|
-
> Please correct anything that is wrong, then say **"confirmed"**.
|
|
135
|
-
```
|
|
136
|
-
|
|
137
|
-
Why it works: The `🔍` signals "I'm showing you what I understood." Bold labels on each bullet make it scannable. The blockquote at the end is the action prompt.
|
|
138
|
-
|
|
139
|
-
---
|
|
140
|
-
|
|
141
|
-
## Anti-Patterns (Never Do This)
|
|
142
|
-
|
|
143
|
-
| Bad | Good | Why |
|
|
144
|
-
|-----|------|-----|
|
|
145
|
-
| Plain paragraph for action prompts | `## 📌 Action Required` header | Paragraphs are invisible |
|
|
146
|
-
| Burying instructions in the middle of text | Action prompt at the END, fenced with `---` | Users scan bottom-first for "what do I do now" |
|
|
147
|
-
| Multiple action items in one paragraph | One action per bold line or bullet | Cognitive overload |
|
|
148
|
-
| Emoji in file-output headers (`# 🔍 Intent`) | Emoji only in terminal display text | Breaks file parsers/TOCs |
|
|
149
|
-
| Wall of unbroken text explaining context | Bullet list with `**bold labels**` | Scannable vs. readable |
|
|
150
|
-
|
|
151
|
-
---
|
|
152
|
-
|
|
153
|
-
## Quick Rules (For Skill Authors)
|
|
154
|
-
|
|
155
|
-
1. **Every phase transition** gets a `---` fenced `##` header with emoji
|
|
156
|
-
2. **Every action prompt** uses `## 📌 Action Required` pattern
|
|
157
|
-
3. **Every status update** starts with an emoji indicator (✅ ⚠️ ❌ ⏳)
|
|
158
|
-
4. **Every user-facing value** (file paths, commands, inputs) uses `inline code`
|
|
159
|
-
5. **Every "what to do next"** instruction uses `**bold**` for the key action word
|
|
160
|
-
6. **Never deliver critical information as a plain paragraph**
|
|
161
|
-
7. **Action prompts go at the END** of the response, not buried in the middle
|
|
1
|
+
# Terminal Output Formatting Standards
|
|
2
|
+
|
|
3
|
+
Claude Code renders markdown via an Ink/React TUI renderer with ANSI escape codes. Plain paragraphs are visually invisible — they blend into the response stream. These rules ensure critical information is scannable.
|
|
4
|
+
|
|
5
|
+
## The Core Problem
|
|
6
|
+
|
|
7
|
+
In a terminal, the user sees a continuous stream of white text. Only these elements create visual differentiation:
|
|
8
|
+
|
|
9
|
+
| Element | Renders As | Use For |
|
|
10
|
+
|---------|-----------|---------|
|
|
11
|
+
| `# H1` / `## H2` | Colored text (red/orange) | Section breaks, banners |
|
|
12
|
+
| `**bold**` | Bold weight | Key terms, labels |
|
|
13
|
+
| `` `inline code` `` | Distinct background/color | Commands, file paths, values |
|
|
14
|
+
| `> blockquote` | Indented with left border | Callouts, important notes |
|
|
15
|
+
| `---` | Horizontal rule | Visual fences/separators |
|
|
16
|
+
| Emojis | Colored glyphs | Status indicators, anchors |
|
|
17
|
+
| Code blocks | Syntax-highlighted box | Code, structured data |
|
|
18
|
+
|
|
19
|
+
**Rule: Anything the user must act on or notice MUST use at least two of these elements. Never deliver critical info as a plain paragraph.**
|
|
20
|
+
|
|
21
|
+
---
|
|
22
|
+
|
|
23
|
+
## Emoji Vocabulary (Standardized)
|
|
24
|
+
|
|
25
|
+
Use consistently so users learn the visual language:
|
|
26
|
+
|
|
27
|
+
| Emoji | Meaning | When to Use |
|
|
28
|
+
|-------|---------|-------------|
|
|
29
|
+
| 🚀 | Workflow start / launch | Welcome banners, phase kickoffs |
|
|
30
|
+
| 📌 | Action required | User needs to do something NOW |
|
|
31
|
+
| ✅ | Success / complete | Task done, tests pass, validation pass |
|
|
32
|
+
| ⚠️ | Warning / attention | Non-blocking but important |
|
|
33
|
+
| ❌ | Error / failure | Something broke, blocking issue |
|
|
34
|
+
| 🔍 | Context / analysis | Showing inferred or extracted info |
|
|
35
|
+
| 📋 | Summary / recap | Phase summaries, status reports |
|
|
36
|
+
| 💡 | Recommendation / tip | AI suggestions, best practices |
|
|
37
|
+
| ⏳ | In progress / waiting | Long-running operations |
|
|
38
|
+
| 🔄 | Iteration / retry | Re-running, looping back |
|
|
39
|
+
|
|
40
|
+
**Do NOT use emojis in markdown headers that will be written to files** (breaks TOCs and parsers). Emojis are for terminal display output only.
|
|
41
|
+
|
|
42
|
+
---
|
|
43
|
+
|
|
44
|
+
## Output Patterns
|
|
45
|
+
|
|
46
|
+
### Pattern 1: Welcome / Phase Banner
|
|
47
|
+
|
|
48
|
+
```markdown
|
|
49
|
+
---
|
|
50
|
+
## 🚀 Welcome to AI-DLC! Starting a new workflow for Group 1B
|
|
51
|
+
|
|
52
|
+
**Workflow:** `group-1b-test-quality` — Test Quality & Traceability
|
|
53
|
+
**Builds on:** Group 1A test infrastructure (confirmed ✅)
|
|
54
|
+
---
|
|
55
|
+
```
|
|
56
|
+
|
|
57
|
+
Why it works: `---` fences create a visual box. `##` renders in color. Emoji grabs the eye. Bold labels + inline code for scannable key-value pairs.
|
|
58
|
+
|
|
59
|
+
---
|
|
60
|
+
|
|
61
|
+
### Pattern 2: Action Required (MOST IMPORTANT)
|
|
62
|
+
|
|
63
|
+
Use this whenever the user needs to do something. This is the pattern that matters most.
|
|
64
|
+
|
|
65
|
+
```markdown
|
|
66
|
+
---
|
|
67
|
+
## 📌 Action Required
|
|
68
|
+
|
|
69
|
+
Fill in the `[Answer]:` tags in the file below, then say **"done"** to continue.
|
|
70
|
+
|
|
71
|
+
> For multi-select questions, provide comma-separated letters (e.g., `A, B, E`).
|
|
72
|
+
|
|
73
|
+
**File:** `aidlc-docs/group-1b-test-quality/inception/intent-questions.md`
|
|
74
|
+
---
|
|
75
|
+
```
|
|
76
|
+
|
|
77
|
+
Why it works: The `📌 Action Required` header is an unmissable colored+emoji signal. The command uses `inline code` and `**bold**`. The secondary instruction lives in a `> blockquote` so it's visually subordinate but still distinct. File path is in code formatting.
|
|
78
|
+
|
|
79
|
+
---
|
|
80
|
+
|
|
81
|
+
### Pattern 3: Status / Progress Update
|
|
82
|
+
|
|
83
|
+
```markdown
|
|
84
|
+
✅ All 3,360 tests pass — Group 1A prerequisite confirmed.
|
|
85
|
+
|
|
86
|
+
⏳ Generating intent questions file...
|
|
87
|
+
```
|
|
88
|
+
|
|
89
|
+
Why it works: Emoji-first lines create a scannable status log. Each line starts with a visual indicator.
|
|
90
|
+
|
|
91
|
+
---
|
|
92
|
+
|
|
93
|
+
### Pattern 4: Validation Summary
|
|
94
|
+
|
|
95
|
+
```markdown
|
|
96
|
+
## 📋 Validation Summary
|
|
97
|
+
|
|
98
|
+
| Check | Status | Detail |
|
|
99
|
+
|-------|--------|--------|
|
|
100
|
+
| Tests pass | ✅ | 3,360 / 3,360 |
|
|
101
|
+
| Prerequisites met | ✅ | Group 1A complete |
|
|
102
|
+
| Manifest valid | ⚠️ | Missing `coverage-config.json` |
|
|
103
|
+
```
|
|
104
|
+
|
|
105
|
+
Why it works: Tables with emoji status create an at-a-glance dashboard. The `##` header separates it from surrounding text.
|
|
106
|
+
|
|
107
|
+
---
|
|
108
|
+
|
|
109
|
+
### Pattern 5: Error / Blocking Issue
|
|
110
|
+
|
|
111
|
+
```markdown
|
|
112
|
+
---
|
|
113
|
+
## ❌ Blocking Issue
|
|
114
|
+
|
|
115
|
+
**What failed:** Contract validation for `UserService.create()`
|
|
116
|
+
**Why:** Missing required field `email` in test fixture at `tests/fixtures/users.ts:42`
|
|
117
|
+
|
|
118
|
+
> Fix the fixture and re-run with **"retry"**.
|
|
119
|
+
---
|
|
120
|
+
```
|
|
121
|
+
|
|
122
|
+
---
|
|
123
|
+
|
|
124
|
+
### Pattern 6: Context Extraction / Inference Review
|
|
125
|
+
|
|
126
|
+
```markdown
|
|
127
|
+
## 🔍 Context Extracted from Your Input
|
|
128
|
+
|
|
129
|
+
- **Problem:** AI-generated tests have documented failure modes...
|
|
130
|
+
- **Primary users:** Olympus OSS users running AI-DLC workflows...
|
|
131
|
+
- **Scope:** Four capabilities — anti-pattern detection, traceability...
|
|
132
|
+
- **Success criteria:** Each capability integrates into the pipeline...
|
|
133
|
+
|
|
134
|
+
> Please correct anything that is wrong, then say **"confirmed"**.
|
|
135
|
+
```
|
|
136
|
+
|
|
137
|
+
Why it works: The `🔍` signals "I'm showing you what I understood." Bold labels on each bullet make it scannable. The blockquote at the end is the action prompt.
|
|
138
|
+
|
|
139
|
+
---
|
|
140
|
+
|
|
141
|
+
## Anti-Patterns (Never Do This)
|
|
142
|
+
|
|
143
|
+
| Bad | Good | Why |
|
|
144
|
+
|-----|------|-----|
|
|
145
|
+
| Plain paragraph for action prompts | `## 📌 Action Required` header | Paragraphs are invisible |
|
|
146
|
+
| Burying instructions in the middle of text | Action prompt at the END, fenced with `---` | Users scan bottom-first for "what do I do now" |
|
|
147
|
+
| Multiple action items in one paragraph | One action per bold line or bullet | Cognitive overload |
|
|
148
|
+
| Emoji in file-output headers (`# 🔍 Intent`) | Emoji only in terminal display text | Breaks file parsers/TOCs |
|
|
149
|
+
| Wall of unbroken text explaining context | Bullet list with `**bold labels**` | Scannable vs. readable |
|
|
150
|
+
|
|
151
|
+
---
|
|
152
|
+
|
|
153
|
+
## Quick Rules (For Skill Authors)
|
|
154
|
+
|
|
155
|
+
1. **Every phase transition** gets a `---` fenced `##` header with emoji
|
|
156
|
+
2. **Every action prompt** uses `## 📌 Action Required` pattern
|
|
157
|
+
3. **Every status update** starts with an emoji indicator (✅ ⚠️ ❌ ⏳)
|
|
158
|
+
4. **Every user-facing value** (file paths, commands, inputs) uses `inline code`
|
|
159
|
+
5. **Every "what to do next"** instruction uses `**bold**` for the key action word
|
|
160
|
+
6. **Never deliver critical information as a plain paragraph**
|
|
161
|
+
7. **Action prompts go at the END** of the response, not buried in the middle
|