powerbi-mcp 0.1.0__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.
Files changed (51) hide show
  1. powerbi_mcp-0.1.0/.claude/settings.json +6 -0
  2. powerbi_mcp-0.1.0/.claude/skills/claude-md-improver/SKILL.md +179 -0
  3. powerbi_mcp-0.1.0/.claude/skills/claude-md-improver/references/quality-criteria.md +109 -0
  4. powerbi_mcp-0.1.0/.claude/skills/claude-md-improver/references/templates.md +253 -0
  5. powerbi_mcp-0.1.0/.claude/skills/claude-md-improver/references/update-guidelines.md +150 -0
  6. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/LICENSE.txt +202 -0
  7. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/SKILL.md +236 -0
  8. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/reference/evaluation.md +602 -0
  9. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/reference/mcp_best_practices.md +249 -0
  10. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/reference/node_mcp_server.md +970 -0
  11. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/reference/python_mcp_server.md +719 -0
  12. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/scripts/connections.py +151 -0
  13. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/scripts/evaluation.py +373 -0
  14. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/scripts/example_evaluation.xml +22 -0
  15. powerbi_mcp-0.1.0/.claude/skills/mcp-builder/scripts/requirements.txt +2 -0
  16. powerbi_mcp-0.1.0/.claude/skills/powerbi-expert/SKILL.md +863 -0
  17. powerbi_mcp-0.1.0/.claude/skills/powerbi-mcp-server-tester/SKILL.md +205 -0
  18. powerbi_mcp-0.1.0/.claude/skills/skill-creator/LICENSE.txt +202 -0
  19. powerbi_mcp-0.1.0/.claude/skills/skill-creator/SKILL.md +356 -0
  20. powerbi_mcp-0.1.0/.claude/skills/skill-creator/references/output-patterns.md +82 -0
  21. powerbi_mcp-0.1.0/.claude/skills/skill-creator/references/workflows.md +28 -0
  22. powerbi_mcp-0.1.0/.claude/skills/skill-creator/scripts/init_skill.py +303 -0
  23. powerbi_mcp-0.1.0/.claude/skills/skill-creator/scripts/package_skill.py +110 -0
  24. powerbi_mcp-0.1.0/.claude/skills/skill-creator/scripts/quick_validate.py +95 -0
  25. powerbi_mcp-0.1.0/.claude/skills/uv-package-manager/SKILL.md +834 -0
  26. powerbi_mcp-0.1.0/.env.example +43 -0
  27. powerbi_mcp-0.1.0/.github/workflows/publish-to-pypi.yml +46 -0
  28. powerbi_mcp-0.1.0/.gitignore +16 -0
  29. powerbi_mcp-0.1.0/.mcp.json +12 -0
  30. powerbi_mcp-0.1.0/.python-version +1 -0
  31. powerbi_mcp-0.1.0/CLAUDE.md +87 -0
  32. powerbi_mcp-0.1.0/DEVELOPER_GUIDE.md +720 -0
  33. powerbi_mcp-0.1.0/LICENSE +21 -0
  34. powerbi_mcp-0.1.0/PKG-INFO +540 -0
  35. powerbi_mcp-0.1.0/README.md +513 -0
  36. powerbi_mcp-0.1.0/opencode.json +12 -0
  37. powerbi_mcp-0.1.0/pyproject.toml +43 -0
  38. powerbi_mcp-0.1.0/src/__init__.py +13 -0
  39. powerbi_mcp-0.1.0/src/auth.py +136 -0
  40. powerbi_mcp-0.1.0/src/client.py +259 -0
  41. powerbi_mcp-0.1.0/src/constants.py +26 -0
  42. powerbi_mcp-0.1.0/src/exceptions.py +96 -0
  43. powerbi_mcp-0.1.0/src/formatters.py +367 -0
  44. powerbi_mcp-0.1.0/src/models.py +85 -0
  45. powerbi_mcp-0.1.0/src/server.py +90 -0
  46. powerbi_mcp-0.1.0/src/tools/__init__.py +15 -0
  47. powerbi_mcp-0.1.0/src/tools/datasets.py +355 -0
  48. powerbi_mcp-0.1.0/src/tools/queries.py +125 -0
  49. powerbi_mcp-0.1.0/src/tools/workspaces.py +185 -0
  50. powerbi_mcp-0.1.0/src/validation.py +40 -0
  51. powerbi_mcp-0.1.0/uv.lock +687 -0
@@ -0,0 +1,6 @@
1
+ {
2
+ "enabledPlugins": {
3
+ "code-simplifier@claude-plugins-official": true,
4
+ "claude-md-management@claude-plugins-official": true
5
+ }
6
+ }
@@ -0,0 +1,179 @@
1
+ ---
2
+ name: claude-md-improver
3
+ description: Audit and improve CLAUDE.md files in repositories. Use when user asks to check, audit, update, improve, or fix CLAUDE.md files. Scans for all CLAUDE.md files, evaluates quality against templates, outputs quality report, then makes targeted updates. Also use when the user mentions "CLAUDE.md maintenance" or "project memory optimization".
4
+ tools: Read, Glob, Grep, Bash, Edit
5
+ ---
6
+
7
+ # CLAUDE.md Improver
8
+
9
+ Audit, evaluate, and improve CLAUDE.md files across a codebase to ensure Claude Code has optimal project context.
10
+
11
+ **This skill can write to CLAUDE.md files.** After presenting a quality report and getting user approval, it updates CLAUDE.md files with targeted improvements.
12
+
13
+ ## Workflow
14
+
15
+ ### Phase 1: Discovery
16
+
17
+ Find all CLAUDE.md files in the repository:
18
+
19
+ ```bash
20
+ find . -name "CLAUDE.md" -o -name ".claude.md" -o -name ".claude.local.md" 2>/dev/null | head -50
21
+ ```
22
+
23
+ **File Types & Locations:**
24
+
25
+ | Type | Location | Purpose |
26
+ |------|----------|---------|
27
+ | Project root | `./CLAUDE.md` | Primary project context (checked into git, shared with team) |
28
+ | Local overrides | `./.claude.local.md` | Personal/local settings (gitignored, not shared) |
29
+ | Global defaults | `~/.claude/CLAUDE.md` | User-wide defaults across all projects |
30
+ | Package-specific | `./packages/*/CLAUDE.md` | Module-level context in monorepos |
31
+ | Subdirectory | Any nested location | Feature/domain-specific context |
32
+
33
+ **Note:** Claude auto-discovers CLAUDE.md files in parent directories, making monorepo setups work automatically.
34
+
35
+ ### Phase 2: Quality Assessment
36
+
37
+ For each CLAUDE.md file, evaluate against quality criteria. See [references/quality-criteria.md](references/quality-criteria.md) for detailed rubrics.
38
+
39
+ **Quick Assessment Checklist:**
40
+
41
+ | Criterion | Weight | Check |
42
+ |-----------|--------|-------|
43
+ | Commands/workflows documented | High | Are build/test/deploy commands present? |
44
+ | Architecture clarity | High | Can Claude understand the codebase structure? |
45
+ | Non-obvious patterns | Medium | Are gotchas and quirks documented? |
46
+ | Conciseness | Medium | No verbose explanations or obvious info? |
47
+ | Currency | High | Does it reflect current codebase state? |
48
+ | Actionability | High | Are instructions executable, not vague? |
49
+
50
+ **Quality Scores:**
51
+ - **A (90-100)**: Comprehensive, current, actionable
52
+ - **B (70-89)**: Good coverage, minor gaps
53
+ - **C (50-69)**: Basic info, missing key sections
54
+ - **D (30-49)**: Sparse or outdated
55
+ - **F (0-29)**: Missing or severely outdated
56
+
57
+ ### Phase 3: Quality Report Output
58
+
59
+ **ALWAYS output the quality report BEFORE making any updates.**
60
+
61
+ Format:
62
+
63
+ ```
64
+ ## CLAUDE.md Quality Report
65
+
66
+ ### Summary
67
+ - Files found: X
68
+ - Average score: X/100
69
+ - Files needing update: X
70
+
71
+ ### File-by-File Assessment
72
+
73
+ #### 1. ./CLAUDE.md (Project Root)
74
+ **Score: XX/100 (Grade: X)**
75
+
76
+ | Criterion | Score | Notes |
77
+ |-----------|-------|-------|
78
+ | Commands/workflows | X/20 | ... |
79
+ | Architecture clarity | X/20 | ... |
80
+ | Non-obvious patterns | X/15 | ... |
81
+ | Conciseness | X/15 | ... |
82
+ | Currency | X/15 | ... |
83
+ | Actionability | X/15 | ... |
84
+
85
+ **Issues:**
86
+ - [List specific problems]
87
+
88
+ **Recommended additions:**
89
+ - [List what should be added]
90
+
91
+ #### 2. ./packages/api/CLAUDE.md (Package-specific)
92
+ ...
93
+ ```
94
+
95
+ ### Phase 4: Targeted Updates
96
+
97
+ After outputting the quality report, ask user for confirmation before updating.
98
+
99
+ **Update Guidelines (Critical):**
100
+
101
+ 1. **Propose targeted additions only** - Focus on genuinely useful info:
102
+ - Commands or workflows discovered during analysis
103
+ - Gotchas or non-obvious patterns found in code
104
+ - Package relationships that weren't clear
105
+ - Testing approaches that work
106
+ - Configuration quirks
107
+
108
+ 2. **Keep it minimal** - Avoid:
109
+ - Restating what's obvious from the code
110
+ - Generic best practices already covered
111
+ - One-off fixes unlikely to recur
112
+ - Verbose explanations when a one-liner suffices
113
+
114
+ 3. **Show diffs** - For each change, show:
115
+ - Which CLAUDE.md file to update
116
+ - The specific addition (as a diff or quoted block)
117
+ - Brief explanation of why this helps future sessions
118
+
119
+ **Diff Format:**
120
+
121
+ ```markdown
122
+ ### Update: ./CLAUDE.md
123
+
124
+ **Why:** Build command was missing, causing confusion about how to run the project.
125
+
126
+ ```diff
127
+ + ## Quick Start
128
+ +
129
+ + ```bash
130
+ + npm install
131
+ + npm run dev # Start development server on port 3000
132
+ + ```
133
+ ```
134
+ ```
135
+
136
+ ### Phase 5: Apply Updates
137
+
138
+ After user approval, apply changes using the Edit tool. Preserve existing content structure.
139
+
140
+ ## Templates
141
+
142
+ See [references/templates.md](references/templates.md) for CLAUDE.md templates by project type.
143
+
144
+ ## Common Issues to Flag
145
+
146
+ 1. **Stale commands**: Build commands that no longer work
147
+ 2. **Missing dependencies**: Required tools not mentioned
148
+ 3. **Outdated architecture**: File structure that's changed
149
+ 4. **Missing environment setup**: Required env vars or config
150
+ 5. **Broken test commands**: Test scripts that have changed
151
+ 6. **Undocumented gotchas**: Non-obvious patterns not captured
152
+
153
+ ## User Tips to Share
154
+
155
+ When presenting recommendations, remind users:
156
+
157
+ - **`#` key shortcut**: During a Claude session, press `#` to have Claude auto-incorporate learnings into CLAUDE.md
158
+ - **Keep it concise**: CLAUDE.md should be human-readable; dense is better than verbose
159
+ - **Actionable commands**: All documented commands should be copy-paste ready
160
+ - **Use `.claude.local.md`**: For personal preferences not shared with team (add to `.gitignore`)
161
+ - **Global defaults**: Put user-wide preferences in `~/.claude/CLAUDE.md`
162
+
163
+ ## What Makes a Great CLAUDE.md
164
+
165
+ **Key principles:**
166
+ - Concise and human-readable
167
+ - Actionable commands that can be copy-pasted
168
+ - Project-specific patterns, not generic advice
169
+ - Non-obvious gotchas and warnings
170
+
171
+ **Recommended sections** (use only what's relevant):
172
+ - Commands (build, test, dev, lint)
173
+ - Architecture (directory structure)
174
+ - Key Files (entry points, config)
175
+ - Code Style (project conventions)
176
+ - Environment (required vars, setup)
177
+ - Testing (commands, patterns)
178
+ - Gotchas (quirks, common mistakes)
179
+ - Workflow (when to do what)
@@ -0,0 +1,109 @@
1
+ # CLAUDE.md Quality Criteria
2
+
3
+ ## Scoring Rubric
4
+
5
+ ### 1. Commands/Workflows (20 points)
6
+
7
+ **20 points**: All essential commands documented with context
8
+ - Build, test, lint, deploy commands present
9
+ - Development workflow clear
10
+ - Common operations documented
11
+
12
+ **15 points**: Most commands present, some missing context
13
+
14
+ **10 points**: Basic commands only, no workflow
15
+
16
+ **5 points**: Few commands, many missing
17
+
18
+ **0 points**: No commands documented
19
+
20
+ ### 2. Architecture Clarity (20 points)
21
+
22
+ **20 points**: Clear codebase map
23
+ - Key directories explained
24
+ - Module relationships documented
25
+ - Entry points identified
26
+ - Data flow described where relevant
27
+
28
+ **15 points**: Good structure overview, minor gaps
29
+
30
+ **10 points**: Basic directory listing only
31
+
32
+ **5 points**: Vague or incomplete
33
+
34
+ **0 points**: No architecture info
35
+
36
+ ### 3. Non-Obvious Patterns (15 points)
37
+
38
+ **15 points**: Gotchas and quirks captured
39
+ - Known issues documented
40
+ - Workarounds explained
41
+ - Edge cases noted
42
+ - "Why we do it this way" for unusual patterns
43
+
44
+ **10 points**: Some patterns documented
45
+
46
+ **5 points**: Minimal pattern documentation
47
+
48
+ **0 points**: No patterns or gotchas
49
+
50
+ ### 4. Conciseness (15 points)
51
+
52
+ **15 points**: Dense, valuable content
53
+ - No filler or obvious info
54
+ - Each line adds value
55
+ - No redundancy with code comments
56
+
57
+ **10 points**: Mostly concise, some padding
58
+
59
+ **5 points**: Verbose in places
60
+
61
+ **0 points**: Mostly filler or restates obvious code
62
+
63
+ ### 5. Currency (15 points)
64
+
65
+ **15 points**: Reflects current codebase
66
+ - Commands work as documented
67
+ - File references accurate
68
+ - Tech stack current
69
+
70
+ **10 points**: Mostly current, minor staleness
71
+
72
+ **5 points**: Several outdated references
73
+
74
+ **0 points**: Severely outdated
75
+
76
+ ### 6. Actionability (15 points)
77
+
78
+ **15 points**: Instructions are executable
79
+ - Commands can be copy-pasted
80
+ - Steps are concrete
81
+ - Paths are real
82
+
83
+ **10 points**: Mostly actionable
84
+
85
+ **5 points**: Some vague instructions
86
+
87
+ **0 points**: Vague or theoretical
88
+
89
+ ## Assessment Process
90
+
91
+ 1. Read the CLAUDE.md file completely
92
+ 2. Cross-reference with actual codebase:
93
+ - Run documented commands (mentally or actually)
94
+ - Check if referenced files exist
95
+ - Verify architecture descriptions
96
+ 3. Score each criterion
97
+ 4. Calculate total and assign grade
98
+ 5. List specific issues found
99
+ 6. Propose concrete improvements
100
+
101
+ ## Red Flags
102
+
103
+ - Commands that would fail (wrong paths, missing deps)
104
+ - References to deleted files/folders
105
+ - Outdated tech versions
106
+ - Copy-paste from templates without customization
107
+ - Generic advice not specific to the project
108
+ - "TODO" items never completed
109
+ - Duplicate info across multiple CLAUDE.md files
@@ -0,0 +1,253 @@
1
+ # CLAUDE.md Templates
2
+
3
+ ## Key Principles
4
+
5
+ - **Concise**: Dense, human-readable content; one line per concept when possible
6
+ - **Actionable**: Commands should be copy-paste ready
7
+ - **Project-specific**: Document patterns unique to this project, not generic advice
8
+ - **Current**: All info should reflect actual codebase state
9
+
10
+ ---
11
+
12
+ ## Recommended Sections
13
+
14
+ Use only the sections relevant to the project. Not all sections are needed.
15
+
16
+ ### Commands
17
+
18
+ Document the essential commands for working with the project.
19
+
20
+ ```markdown
21
+ ## Commands
22
+
23
+ | Command | Description |
24
+ |---------|-------------|
25
+ | `<install command>` | Install dependencies |
26
+ | `<dev command>` | Start development server |
27
+ | `<build command>` | Production build |
28
+ | `<test command>` | Run tests |
29
+ | `<lint command>` | Lint/format code |
30
+ ```
31
+
32
+ ### Architecture
33
+
34
+ Describe the project structure so Claude understands where things live.
35
+
36
+ ```markdown
37
+ ## Architecture
38
+
39
+ ```
40
+ <root>/
41
+ <dir>/ # <purpose>
42
+ <dir>/ # <purpose>
43
+ <dir>/ # <purpose>
44
+ ```
45
+ ```
46
+
47
+ ### Key Files
48
+
49
+ List important files that Claude should know about.
50
+
51
+ ```markdown
52
+ ## Key Files
53
+
54
+ - `<path>` - <purpose>
55
+ - `<path>` - <purpose>
56
+ ```
57
+
58
+ ### Code Style
59
+
60
+ Document project-specific coding conventions.
61
+
62
+ ```markdown
63
+ ## Code Style
64
+
65
+ - <convention>
66
+ - <convention>
67
+ - <preference over alternative>
68
+ ```
69
+
70
+ ### Environment
71
+
72
+ Document required environment variables and setup.
73
+
74
+ ```markdown
75
+ ## Environment
76
+
77
+ Required:
78
+ - `<VAR_NAME>` - <purpose>
79
+ - `<VAR_NAME>` - <purpose>
80
+
81
+ Setup:
82
+ - <setup step>
83
+ ```
84
+
85
+ ### Testing
86
+
87
+ Document testing approach and commands.
88
+
89
+ ```markdown
90
+ ## Testing
91
+
92
+ - `<test command>` - <what it tests>
93
+ - <testing convention or pattern>
94
+ ```
95
+
96
+ ### Gotchas
97
+
98
+ Document non-obvious patterns, quirks, and warnings.
99
+
100
+ ```markdown
101
+ ## Gotchas
102
+
103
+ - <non-obvious thing that causes issues>
104
+ - <ordering dependency or prerequisite>
105
+ - <common mistake to avoid>
106
+ ```
107
+
108
+ ### Workflow
109
+
110
+ Document development workflow patterns.
111
+
112
+ ```markdown
113
+ ## Workflow
114
+
115
+ - <when to do X>
116
+ - <preferred approach for Y>
117
+ ```
118
+
119
+ ---
120
+
121
+ ## Template: Project Root (Minimal)
122
+
123
+ ```markdown
124
+ # <Project Name>
125
+
126
+ <One-line description>
127
+
128
+ ## Commands
129
+
130
+ | Command | Description |
131
+ |---------|-------------|
132
+ | `<command>` | <description> |
133
+
134
+ ## Architecture
135
+
136
+ ```
137
+ <structure>
138
+ ```
139
+
140
+ ## Gotchas
141
+
142
+ - <gotcha>
143
+ ```
144
+
145
+ ---
146
+
147
+ ## Template: Project Root (Comprehensive)
148
+
149
+ ```markdown
150
+ # <Project Name>
151
+
152
+ <One-line description>
153
+
154
+ ## Commands
155
+
156
+ | Command | Description |
157
+ |---------|-------------|
158
+ | `<command>` | <description> |
159
+
160
+ ## Architecture
161
+
162
+ ```
163
+ <structure with descriptions>
164
+ ```
165
+
166
+ ## Key Files
167
+
168
+ - `<path>` - <purpose>
169
+
170
+ ## Code Style
171
+
172
+ - <convention>
173
+
174
+ ## Environment
175
+
176
+ - `<VAR>` - <purpose>
177
+
178
+ ## Testing
179
+
180
+ - `<command>` - <scope>
181
+
182
+ ## Gotchas
183
+
184
+ - <gotcha>
185
+ ```
186
+
187
+ ---
188
+
189
+ ## Template: Package/Module
190
+
191
+ For packages within a monorepo or distinct modules.
192
+
193
+ ```markdown
194
+ # <Package Name>
195
+
196
+ <Purpose of this package>
197
+
198
+ ## Usage
199
+
200
+ ```
201
+ <import/usage example>
202
+ ```
203
+
204
+ ## Key Exports
205
+
206
+ - `<export>` - <purpose>
207
+
208
+ ## Dependencies
209
+
210
+ - `<dependency>` - <why needed>
211
+
212
+ ## Notes
213
+
214
+ - <important note>
215
+ ```
216
+
217
+ ---
218
+
219
+ ## Template: Monorepo Root
220
+
221
+ ```markdown
222
+ # <Monorepo Name>
223
+
224
+ <Description>
225
+
226
+ ## Packages
227
+
228
+ | Package | Description | Path |
229
+ |---------|-------------|------|
230
+ | `<name>` | <purpose> | `<path>` |
231
+
232
+ ## Commands
233
+
234
+ | Command | Description |
235
+ |---------|-------------|
236
+ | `<command>` | <description> |
237
+
238
+ ## Cross-Package Patterns
239
+
240
+ - <shared pattern>
241
+ - <generation/sync pattern>
242
+ ```
243
+
244
+ ---
245
+
246
+ ## Update Principles
247
+
248
+ When updating any CLAUDE.md:
249
+
250
+ 1. **Be specific**: Use actual file paths, real commands from this project
251
+ 2. **Be current**: Verify info against the actual codebase
252
+ 3. **Be brief**: One line per concept when possible
253
+ 4. **Be useful**: Would this help a new Claude session understand the project?
@@ -0,0 +1,150 @@
1
+ # CLAUDE.md Update Guidelines
2
+
3
+ ## Core Principle
4
+
5
+ Only add information that will genuinely help future Claude sessions. The context window is precious - every line must earn its place.
6
+
7
+ ## What TO Add
8
+
9
+ ### 1. Commands/Workflows Discovered
10
+
11
+ ```markdown
12
+ ## Build
13
+
14
+ `npm run build:prod` - Full production build with optimization
15
+ `npm run build:dev` - Fast dev build (no minification)
16
+ ```
17
+
18
+ Why: Saves future sessions from discovering these again.
19
+
20
+ ### 2. Gotchas and Non-Obvious Patterns
21
+
22
+ ```markdown
23
+ ## Gotchas
24
+
25
+ - Tests must run sequentially (`--runInBand`) due to shared DB state
26
+ - `yarn.lock` is authoritative; delete `node_modules` if deps mismatch
27
+ ```
28
+
29
+ Why: Prevents repeating debugging sessions.
30
+
31
+ ### 3. Package Relationships
32
+
33
+ ```markdown
34
+ ## Dependencies
35
+
36
+ The `auth` module depends on `crypto` being initialized first.
37
+ Import order matters in `src/bootstrap.ts`.
38
+ ```
39
+
40
+ Why: Architecture knowledge that isn't obvious from code.
41
+
42
+ ### 4. Testing Approaches That Worked
43
+
44
+ ```markdown
45
+ ## Testing
46
+
47
+ For API endpoints: Use `supertest` with the test helper in `tests/setup.ts`
48
+ Mocking: Factory functions in `tests/factories/` (not inline mocks)
49
+ ```
50
+
51
+ Why: Establishes patterns that work.
52
+
53
+ ### 5. Configuration Quirks
54
+
55
+ ```markdown
56
+ ## Config
57
+
58
+ - `NEXT_PUBLIC_*` vars must be set at build time, not runtime
59
+ - Redis connection requires `?family=0` suffix for IPv6
60
+ ```
61
+
62
+ Why: Environment-specific knowledge.
63
+
64
+ ## What NOT to Add
65
+
66
+ ### 1. Obvious Code Info
67
+
68
+ Bad:
69
+ ```markdown
70
+ The `UserService` class handles user operations.
71
+ ```
72
+
73
+ The class name already tells us this.
74
+
75
+ ### 2. Generic Best Practices
76
+
77
+ Bad:
78
+ ```markdown
79
+ Always write tests for new features.
80
+ Use meaningful variable names.
81
+ ```
82
+
83
+ This is universal advice, not project-specific.
84
+
85
+ ### 3. One-Off Fixes
86
+
87
+ Bad:
88
+ ```markdown
89
+ We fixed a bug in commit abc123 where the login button didn't work.
90
+ ```
91
+
92
+ Won't recur; clutters the file.
93
+
94
+ ### 4. Verbose Explanations
95
+
96
+ Bad:
97
+ ```markdown
98
+ The authentication system uses JWT tokens. JWT (JSON Web Tokens) are
99
+ an open standard (RFC 7519) that defines a compact and self-contained
100
+ way for securely transmitting information between parties as a JSON
101
+ object. In our implementation, we use the HS256 algorithm which...
102
+ ```
103
+
104
+ Good:
105
+ ```markdown
106
+ Auth: JWT with HS256, tokens in `Authorization: Bearer <token>` header.
107
+ ```
108
+
109
+ ## Diff Format for Updates
110
+
111
+ For each suggested change:
112
+
113
+ ### 1. Identify the File
114
+
115
+ ```
116
+ File: ./CLAUDE.md
117
+ Section: Commands (new section after ## Architecture)
118
+ ```
119
+
120
+ ### 2. Show the Change
121
+
122
+ ```diff
123
+ ## Architecture
124
+ ...
125
+
126
+ +## Commands
127
+ +
128
+ +| Command | Purpose |
129
+ +|---------|---------|
130
+ +| `npm run dev` | Dev server with HMR |
131
+ +| `npm run build` | Production build |
132
+ +| `npm test` | Run test suite |
133
+ ```
134
+
135
+ ### 3. Explain Why
136
+
137
+ > **Why this helps:** The build commands weren't documented, causing
138
+ > confusion about how to run the project. This saves future sessions
139
+ > from needing to inspect `package.json`.
140
+
141
+ ## Validation Checklist
142
+
143
+ Before finalizing an update, verify:
144
+
145
+ - [ ] Each addition is project-specific
146
+ - [ ] No generic advice or obvious info
147
+ - [ ] Commands are tested and work
148
+ - [ ] File paths are accurate
149
+ - [ ] Would a new Claude session find this helpful?
150
+ - [ ] Is this the most concise way to express the info?