@undeemed/get-shit-done-codex 1.6.5

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 (79) hide show
  1. package/AGENTS.md +183 -0
  2. package/LICENSE +21 -0
  3. package/README.md +179 -0
  4. package/bin/install.js +214 -0
  5. package/commands/gsd/add-phase.md +207 -0
  6. package/commands/gsd/add-todo.md +182 -0
  7. package/commands/gsd/audit-milestone.md +258 -0
  8. package/commands/gsd/check-todos.md +217 -0
  9. package/commands/gsd/complete-milestone.md +136 -0
  10. package/commands/gsd/debug.md +149 -0
  11. package/commands/gsd/discuss-phase.md +80 -0
  12. package/commands/gsd/execute-phase.md +304 -0
  13. package/commands/gsd/help.md +383 -0
  14. package/commands/gsd/insert-phase.md +227 -0
  15. package/commands/gsd/list-phase-assumptions.md +50 -0
  16. package/commands/gsd/map-codebase.md +71 -0
  17. package/commands/gsd/new-milestone.md +717 -0
  18. package/commands/gsd/new-project.md +896 -0
  19. package/commands/gsd/pause-work.md +123 -0
  20. package/commands/gsd/plan-milestone-gaps.md +284 -0
  21. package/commands/gsd/plan-phase.md +475 -0
  22. package/commands/gsd/progress.md +356 -0
  23. package/commands/gsd/remove-phase.md +338 -0
  24. package/commands/gsd/research-phase.md +180 -0
  25. package/commands/gsd/resume-work.md +40 -0
  26. package/commands/gsd/update.md +172 -0
  27. package/commands/gsd/verify-work.md +219 -0
  28. package/commands/gsd/whats-new.md +124 -0
  29. package/get-shit-done/references/checkpoints.md +788 -0
  30. package/get-shit-done/references/continuation-format.md +249 -0
  31. package/get-shit-done/references/git-integration.md +254 -0
  32. package/get-shit-done/references/questioning.md +141 -0
  33. package/get-shit-done/references/tdd.md +263 -0
  34. package/get-shit-done/references/ui-brand.md +160 -0
  35. package/get-shit-done/references/verification-patterns.md +595 -0
  36. package/get-shit-done/templates/DEBUG.md +159 -0
  37. package/get-shit-done/templates/UAT.md +247 -0
  38. package/get-shit-done/templates/codebase/architecture.md +255 -0
  39. package/get-shit-done/templates/codebase/concerns.md +310 -0
  40. package/get-shit-done/templates/codebase/conventions.md +307 -0
  41. package/get-shit-done/templates/codebase/integrations.md +280 -0
  42. package/get-shit-done/templates/codebase/stack.md +186 -0
  43. package/get-shit-done/templates/codebase/structure.md +285 -0
  44. package/get-shit-done/templates/codebase/testing.md +480 -0
  45. package/get-shit-done/templates/config.json +26 -0
  46. package/get-shit-done/templates/context.md +291 -0
  47. package/get-shit-done/templates/continue-here.md +78 -0
  48. package/get-shit-done/templates/debug-subagent-prompt.md +91 -0
  49. package/get-shit-done/templates/discovery.md +146 -0
  50. package/get-shit-done/templates/milestone-archive.md +123 -0
  51. package/get-shit-done/templates/milestone.md +115 -0
  52. package/get-shit-done/templates/phase-prompt.md +576 -0
  53. package/get-shit-done/templates/planner-subagent-prompt.md +117 -0
  54. package/get-shit-done/templates/project.md +184 -0
  55. package/get-shit-done/templates/requirements.md +231 -0
  56. package/get-shit-done/templates/research-project/ARCHITECTURE.md +204 -0
  57. package/get-shit-done/templates/research-project/FEATURES.md +147 -0
  58. package/get-shit-done/templates/research-project/PITFALLS.md +200 -0
  59. package/get-shit-done/templates/research-project/STACK.md +120 -0
  60. package/get-shit-done/templates/research-project/SUMMARY.md +170 -0
  61. package/get-shit-done/templates/research.md +529 -0
  62. package/get-shit-done/templates/roadmap.md +202 -0
  63. package/get-shit-done/templates/state.md +206 -0
  64. package/get-shit-done/templates/summary.md +269 -0
  65. package/get-shit-done/templates/user-setup.md +323 -0
  66. package/get-shit-done/templates/verification-report.md +322 -0
  67. package/get-shit-done/workflows/complete-milestone.md +750 -0
  68. package/get-shit-done/workflows/diagnose-issues.md +233 -0
  69. package/get-shit-done/workflows/discovery-phase.md +293 -0
  70. package/get-shit-done/workflows/discuss-phase.md +422 -0
  71. package/get-shit-done/workflows/execute-phase.md +552 -0
  72. package/get-shit-done/workflows/execute-plan.md +1831 -0
  73. package/get-shit-done/workflows/list-phase-assumptions.md +178 -0
  74. package/get-shit-done/workflows/map-codebase.md +289 -0
  75. package/get-shit-done/workflows/resume-project.md +311 -0
  76. package/get-shit-done/workflows/transition.md +564 -0
  77. package/get-shit-done/workflows/verify-phase.md +629 -0
  78. package/get-shit-done/workflows/verify-work.md +563 -0
  79. package/package.json +39 -0
@@ -0,0 +1,249 @@
1
+ # Continuation Format
2
+
3
+ Standard format for presenting next steps after completing a command or workflow.
4
+
5
+ ## Core Structure
6
+
7
+ ```
8
+ ---
9
+
10
+ ## ▶ Next Up
11
+
12
+ **{identifier}: {name}** — {one-line description}
13
+
14
+ `{command to copy-paste}`
15
+
16
+ <sub>`/clear` first → fresh context window</sub>
17
+
18
+ ---
19
+
20
+ **Also available:**
21
+ - `{alternative option 1}` — description
22
+ - `{alternative option 2}` — description
23
+
24
+ ---
25
+ ```
26
+
27
+ ## Format Rules
28
+
29
+ 1. **Always show what it is** — name + description, never just a command path
30
+ 2. **Pull context from source** — ROADMAP.md for phases, PLAN.md `<objective>` for plans
31
+ 3. **Command in inline code** — backticks, easy to copy-paste, renders as clickable link
32
+ 4. **`/clear` explanation** — always include, keeps it concise but explains why
33
+ 5. **"Also available" not "Other options"** — sounds more app-like
34
+ 6. **Visual separators** — `---` above and below to make it stand out
35
+
36
+ ## Variants
37
+
38
+ ### Execute Next Plan
39
+
40
+ ```
41
+ ---
42
+
43
+ ## ▶ Next Up
44
+
45
+ **02-03: Refresh Token Rotation** — Add /api/auth/refresh with sliding expiry
46
+
47
+ `/gsd:execute-phase 2`
48
+
49
+ <sub>`/clear` first → fresh context window</sub>
50
+
51
+ ---
52
+
53
+ **Also available:**
54
+ - Review plan before executing
55
+ - `/gsd:list-phase-assumptions 2` — check assumptions
56
+
57
+ ---
58
+ ```
59
+
60
+ ### Execute Final Plan in Phase
61
+
62
+ Add note that this is the last plan and what comes after:
63
+
64
+ ```
65
+ ---
66
+
67
+ ## ▶ Next Up
68
+
69
+ **02-03: Refresh Token Rotation** — Add /api/auth/refresh with sliding expiry
70
+ <sub>Final plan in Phase 2</sub>
71
+
72
+ `/gsd:execute-phase 2`
73
+
74
+ <sub>`/clear` first → fresh context window</sub>
75
+
76
+ ---
77
+
78
+ **After this completes:**
79
+ - Phase 2 → Phase 3 transition
80
+ - Next: **Phase 3: Core Features** — User dashboard and settings
81
+
82
+ ---
83
+ ```
84
+
85
+ ### Plan a Phase
86
+
87
+ ```
88
+ ---
89
+
90
+ ## ▶ Next Up
91
+
92
+ **Phase 2: Authentication** — JWT login flow with refresh tokens
93
+
94
+ `/gsd:plan-phase 2`
95
+
96
+ <sub>`/clear` first → fresh context window</sub>
97
+
98
+ ---
99
+
100
+ **Also available:**
101
+ - `/gsd:discuss-phase 2` — gather context first
102
+ - `/gsd:research-phase 2` — investigate unknowns
103
+ - Review roadmap
104
+
105
+ ---
106
+ ```
107
+
108
+ ### Phase Complete, Ready for Next
109
+
110
+ Show completion status before next action:
111
+
112
+ ```
113
+ ---
114
+
115
+ ## ✓ Phase 2 Complete
116
+
117
+ 3/3 plans executed
118
+
119
+ ## ▶ Next Up
120
+
121
+ **Phase 3: Core Features** — User dashboard, settings, and data export
122
+
123
+ `/gsd:plan-phase 3`
124
+
125
+ <sub>`/clear` first → fresh context window</sub>
126
+
127
+ ---
128
+
129
+ **Also available:**
130
+ - `/gsd:discuss-phase 3` — gather context first
131
+ - `/gsd:research-phase 3` — investigate unknowns
132
+ - Review what Phase 2 built
133
+
134
+ ---
135
+ ```
136
+
137
+ ### Multiple Equal Options
138
+
139
+ When there's no clear primary action:
140
+
141
+ ```
142
+ ---
143
+
144
+ ## ▶ Next Up
145
+
146
+ **Phase 3: Core Features** — User dashboard, settings, and data export
147
+
148
+ **To plan directly:** `/gsd:plan-phase 3`
149
+
150
+ **To discuss context first:** `/gsd:discuss-phase 3`
151
+
152
+ **To research unknowns:** `/gsd:research-phase 3`
153
+
154
+ <sub>`/clear` first → fresh context window</sub>
155
+
156
+ ---
157
+ ```
158
+
159
+ ### Milestone Complete
160
+
161
+ ```
162
+ ---
163
+
164
+ ## 🎉 Milestone v1.0 Complete
165
+
166
+ All 4 phases shipped
167
+
168
+ ## ▶ Next Up
169
+
170
+ **Start v1.1** — questioning → research → requirements → roadmap
171
+
172
+ `/gsd:new-milestone`
173
+
174
+ <sub>`/clear` first → fresh context window</sub>
175
+
176
+ ---
177
+ ```
178
+
179
+ ## Pulling Context
180
+
181
+ ### For phases (from ROADMAP.md):
182
+
183
+ ```markdown
184
+ ### Phase 2: Authentication
185
+ **Goal**: JWT login flow with refresh tokens
186
+ ```
187
+
188
+ Extract: `**Phase 2: Authentication** — JWT login flow with refresh tokens`
189
+
190
+ ### For plans (from ROADMAP.md):
191
+
192
+ ```markdown
193
+ Plans:
194
+ - [ ] 02-03: Add refresh token rotation
195
+ ```
196
+
197
+ Or from PLAN.md `<objective>`:
198
+
199
+ ```xml
200
+ <objective>
201
+ Add refresh token rotation with sliding expiry window.
202
+
203
+ Purpose: Extend session lifetime without compromising security.
204
+ </objective>
205
+ ```
206
+
207
+ Extract: `**02-03: Refresh Token Rotation** — Add /api/auth/refresh with sliding expiry`
208
+
209
+ ## Anti-Patterns
210
+
211
+ ### Don't: Command-only (no context)
212
+
213
+ ```
214
+ ## To Continue
215
+
216
+ Run `/clear`, then paste:
217
+ /gsd:execute-phase 2
218
+ ```
219
+
220
+ User has no idea what 02-03 is about.
221
+
222
+ ### Don't: Missing /clear explanation
223
+
224
+ ```
225
+ `/gsd:plan-phase 3`
226
+
227
+ Run /clear first.
228
+ ```
229
+
230
+ Doesn't explain why. User might skip it.
231
+
232
+ ### Don't: "Other options" language
233
+
234
+ ```
235
+ Other options:
236
+ - Review roadmap
237
+ ```
238
+
239
+ Sounds like an afterthought. Use "Also available:" instead.
240
+
241
+ ### Don't: Fenced code blocks for commands
242
+
243
+ ```
244
+ ```
245
+ /gsd:plan-phase 3
246
+ ```
247
+ ```
248
+
249
+ Fenced blocks inside templates create nesting ambiguity. Use inline backticks instead.
@@ -0,0 +1,254 @@
1
+ <overview>
2
+ Git integration for GSD framework.
3
+ </overview>
4
+
5
+ <core_principle>
6
+
7
+ **Commit outcomes, not process.**
8
+
9
+ The git log should read like a changelog of what shipped, not a diary of planning activity.
10
+ </core_principle>
11
+
12
+ <commit_points>
13
+
14
+ | Event | Commit? | Why |
15
+ | ----------------------- | ------- | ------------------------------------------------ |
16
+ | BRIEF + ROADMAP created | YES | Project initialization |
17
+ | PLAN.md created | NO | Intermediate - commit with plan completion |
18
+ | RESEARCH.md created | NO | Intermediate |
19
+ | DISCOVERY.md created | NO | Intermediate |
20
+ | **Task completed** | YES | Atomic unit of work (1 commit per task) |
21
+ | **Plan completed** | YES | Metadata commit (SUMMARY + STATE + ROADMAP) |
22
+ | Handoff created | YES | WIP state preserved |
23
+
24
+ </commit_points>
25
+
26
+ <git_check>
27
+
28
+ ```bash
29
+ [ -d .git ] && echo "GIT_EXISTS" || echo "NO_GIT"
30
+ ```
31
+
32
+ If NO_GIT: Run `git init` silently. GSD projects always get their own repo.
33
+ </git_check>
34
+
35
+ <commit_formats>
36
+
37
+ <format name="initialization">
38
+ ## Project Initialization (brief + roadmap together)
39
+
40
+ ```
41
+ docs: initialize [project-name] ([N] phases)
42
+
43
+ [One-liner from PROJECT.md]
44
+
45
+ Phases:
46
+ 1. [phase-name]: [goal]
47
+ 2. [phase-name]: [goal]
48
+ 3. [phase-name]: [goal]
49
+ ```
50
+
51
+ What to commit:
52
+
53
+ ```bash
54
+ git add .planning/
55
+ git commit
56
+ ```
57
+
58
+ </format>
59
+
60
+ <format name="task-completion">
61
+ ## Task Completion (During Plan Execution)
62
+
63
+ Each task gets its own commit immediately after completion.
64
+
65
+ ```
66
+ {type}({phase}-{plan}): {task-name}
67
+
68
+ - [Key change 1]
69
+ - [Key change 2]
70
+ - [Key change 3]
71
+ ```
72
+
73
+ **Commit types:**
74
+ - `feat` - New feature/functionality
75
+ - `fix` - Bug fix
76
+ - `test` - Test-only (TDD RED phase)
77
+ - `refactor` - Code cleanup (TDD REFACTOR phase)
78
+ - `perf` - Performance improvement
79
+ - `chore` - Dependencies, config, tooling
80
+
81
+ **Examples:**
82
+
83
+ ```bash
84
+ # Standard task
85
+ git add src/api/auth.ts src/types/user.ts
86
+ git commit -m "feat(08-02): create user registration endpoint
87
+
88
+ - POST /auth/register validates email and password
89
+ - Checks for duplicate users
90
+ - Returns JWT token on success
91
+ "
92
+
93
+ # TDD task - RED phase
94
+ git add src/__tests__/jwt.test.ts
95
+ git commit -m "test(07-02): add failing test for JWT generation
96
+
97
+ - Tests token contains user ID claim
98
+ - Tests token expires in 1 hour
99
+ - Tests signature verification
100
+ "
101
+
102
+ # TDD task - GREEN phase
103
+ git add src/utils/jwt.ts
104
+ git commit -m "feat(07-02): implement JWT generation
105
+
106
+ - Uses jose library for signing
107
+ - Includes user ID and expiry claims
108
+ - Signs with HS256 algorithm
109
+ "
110
+ ```
111
+
112
+ </format>
113
+
114
+ <format name="plan-completion">
115
+ ## Plan Completion (After All Tasks Done)
116
+
117
+ After all tasks committed, one final metadata commit captures plan completion.
118
+
119
+ ```
120
+ docs({phase}-{plan}): complete [plan-name] plan
121
+
122
+ Tasks completed: [N]/[N]
123
+ - [Task 1 name]
124
+ - [Task 2 name]
125
+ - [Task 3 name]
126
+
127
+ SUMMARY: .planning/phases/XX-name/{phase}-{plan}-SUMMARY.md
128
+ ```
129
+
130
+ What to commit:
131
+
132
+ ```bash
133
+ git add .planning/phases/XX-name/{phase}-{plan}-PLAN.md
134
+ git add .planning/phases/XX-name/{phase}-{plan}-SUMMARY.md
135
+ git add .planning/STATE.md
136
+ git add .planning/ROADMAP.md
137
+ git commit
138
+ ```
139
+
140
+ **Note:** Code files NOT included - already committed per-task.
141
+
142
+ </format>
143
+
144
+ <format name="handoff">
145
+ ## Handoff (WIP)
146
+
147
+ ```
148
+ wip: [phase-name] paused at task [X]/[Y]
149
+
150
+ Current: [task name]
151
+ [If blocked:] Blocked: [reason]
152
+ ```
153
+
154
+ What to commit:
155
+
156
+ ```bash
157
+ git add .planning/
158
+ git commit
159
+ ```
160
+
161
+ </format>
162
+ </commit_formats>
163
+
164
+ <example_log>
165
+
166
+ **Old approach (per-plan commits):**
167
+ ```
168
+ a7f2d1 feat(checkout): Stripe payments with webhook verification
169
+ 3e9c4b feat(products): catalog with search, filters, and pagination
170
+ 8a1b2c feat(auth): JWT with refresh rotation using jose
171
+ 5c3d7e feat(foundation): Next.js 15 + Prisma + Tailwind scaffold
172
+ 2f4a8d docs: initialize ecommerce-app (5 phases)
173
+ ```
174
+
175
+ **New approach (per-task commits):**
176
+ ```
177
+ # Phase 04 - Checkout
178
+ 1a2b3c docs(04-01): complete checkout flow plan
179
+ 4d5e6f feat(04-01): add webhook signature verification
180
+ 7g8h9i feat(04-01): implement payment session creation
181
+ 0j1k2l feat(04-01): create checkout page component
182
+
183
+ # Phase 03 - Products
184
+ 3m4n5o docs(03-02): complete product listing plan
185
+ 6p7q8r feat(03-02): add pagination controls
186
+ 9s0t1u feat(03-02): implement search and filters
187
+ 2v3w4x feat(03-01): create product catalog schema
188
+
189
+ # Phase 02 - Auth
190
+ 5y6z7a docs(02-02): complete token refresh plan
191
+ 8b9c0d feat(02-02): implement refresh token rotation
192
+ 1e2f3g test(02-02): add failing test for token refresh
193
+ 4h5i6j docs(02-01): complete JWT setup plan
194
+ 7k8l9m feat(02-01): add JWT generation and validation
195
+ 0n1o2p chore(02-01): install jose library
196
+
197
+ # Phase 01 - Foundation
198
+ 3q4r5s docs(01-01): complete scaffold plan
199
+ 6t7u8v feat(01-01): configure Tailwind and globals
200
+ 9w0x1y feat(01-01): set up Prisma with database
201
+ 2z3a4b feat(01-01): create Next.js 15 project
202
+
203
+ # Initialization
204
+ 5c6d7e docs: initialize ecommerce-app (5 phases)
205
+ ```
206
+
207
+ Each plan produces 2-4 commits (tasks + metadata). Clear, granular, bisectable.
208
+
209
+ </example_log>
210
+
211
+ <anti_patterns>
212
+
213
+ **Still don't commit (intermediate artifacts):**
214
+ - PLAN.md creation (commit with plan completion)
215
+ - RESEARCH.md (intermediate)
216
+ - DISCOVERY.md (intermediate)
217
+ - Minor planning tweaks
218
+ - "Fixed typo in roadmap"
219
+
220
+ **Do commit (outcomes):**
221
+ - Each task completion (feat/fix/test/refactor)
222
+ - Plan completion metadata (docs)
223
+ - Project initialization (docs)
224
+
225
+ **Key principle:** Commit working code and shipped outcomes, not planning process.
226
+
227
+ </anti_patterns>
228
+
229
+ <commit_strategy_rationale>
230
+
231
+ ## Why Per-Task Commits?
232
+
233
+ **Context engineering for AI:**
234
+ - Git history becomes primary context source for future Claude sessions
235
+ - `git log --grep="{phase}-{plan}"` shows all work for a plan
236
+ - `git diff <hash>^..<hash>` shows exact changes per task
237
+ - Less reliance on parsing SUMMARY.md = more context for actual work
238
+
239
+ **Failure recovery:**
240
+ - Task 1 committed ✅, Task 2 failed ❌
241
+ - Claude in next session: sees task 1 complete, can retry task 2
242
+ - Can `git reset --hard` to last successful task
243
+
244
+ **Debugging:**
245
+ - `git bisect` finds exact failing task, not just failing plan
246
+ - `git blame` traces line to specific task context
247
+ - Each commit is independently revertable
248
+
249
+ **Observability:**
250
+ - Solo developer + Claude workflow benefits from granular attribution
251
+ - Atomic commits are git best practice
252
+ - "Commit noise" irrelevant when consumer is Claude, not humans
253
+
254
+ </commit_strategy_rationale>
@@ -0,0 +1,141 @@
1
+ <questioning_guide>
2
+
3
+ Project initialization is dream extraction, not requirements gathering. You're helping the user discover and articulate what they want to build. This isn't a contract negotiation — it's collaborative thinking.
4
+
5
+ <philosophy>
6
+
7
+ **You are a thinking partner, not an interviewer.**
8
+
9
+ The user often has a fuzzy idea. Your job is to help them sharpen it. Ask questions that make them think "oh, I hadn't considered that" or "yes, that's exactly what I mean."
10
+
11
+ Don't interrogate. Collaborate. Don't follow a script. Follow the thread.
12
+
13
+ </philosophy>
14
+
15
+ <the_goal>
16
+
17
+ By the end of questioning, you need enough clarity to write a PROJECT.md that downstream phases can act on:
18
+
19
+ - **Research** needs: what domain to research, what the user already knows, what unknowns exist
20
+ - **Requirements** needs: clear enough vision to scope v1 features
21
+ - **Roadmap** needs: clear enough vision to decompose into phases, what "done" looks like
22
+ - **plan-phase** needs: specific requirements to break into tasks, context for implementation choices
23
+ - **execute-phase** needs: success criteria to verify against, the "why" behind requirements
24
+
25
+ A vague PROJECT.md forces every downstream phase to guess. The cost compounds.
26
+
27
+ </the_goal>
28
+
29
+ <how_to_question>
30
+
31
+ **Start open.** Let them dump their mental model. Don't interrupt with structure.
32
+
33
+ **Follow energy.** Whatever they emphasized, dig into that. What excited them? What problem sparked this?
34
+
35
+ **Challenge vagueness.** Never accept fuzzy answers. "Good" means what? "Users" means who? "Simple" means how?
36
+
37
+ **Make the abstract concrete.** "Walk me through using this." "What does that actually look like?"
38
+
39
+ **Clarify ambiguity.** "When you say Z, do you mean A or B?" "You mentioned X — tell me more."
40
+
41
+ **Know when to stop.** When you understand what they want, why they want it, who it's for, and what done looks like — offer to proceed.
42
+
43
+ </how_to_question>
44
+
45
+ <question_types>
46
+
47
+ Use these as inspiration, not a checklist. Pick what's relevant to the thread.
48
+
49
+ **Motivation — why this exists:**
50
+ - "What prompted this?"
51
+ - "What are you doing today that this replaces?"
52
+ - "What would you do if this existed?"
53
+
54
+ **Concreteness — what it actually is:**
55
+ - "Walk me through using this"
56
+ - "You said X — what does that actually look like?"
57
+ - "Give me an example"
58
+
59
+ **Clarification — what they mean:**
60
+ - "When you say Z, do you mean A or B?"
61
+ - "You mentioned X — tell me more about that"
62
+
63
+ **Success — how you'll know it's working:**
64
+ - "How will you know this is working?"
65
+ - "What does done look like?"
66
+
67
+ </question_types>
68
+
69
+ <using_askuserquestion>
70
+
71
+ Use AskUserQuestion to help users think by presenting concrete options to react to.
72
+
73
+ **Good options:**
74
+ - Interpretations of what they might mean
75
+ - Specific examples to confirm or deny
76
+ - Concrete choices that reveal priorities
77
+
78
+ **Bad options:**
79
+ - Generic categories ("Technical", "Business", "Other")
80
+ - Leading options that presume an answer
81
+ - Too many options (2-4 is ideal)
82
+
83
+ **Example — vague answer:**
84
+ User says "it should be fast"
85
+
86
+ - header: "Fast"
87
+ - question: "Fast how?"
88
+ - options: ["Sub-second response", "Handles large datasets", "Quick to build", "Let me explain"]
89
+
90
+ **Example — following a thread:**
91
+ User mentions "frustrated with current tools"
92
+
93
+ - header: "Frustration"
94
+ - question: "What specifically frustrates you?"
95
+ - options: ["Too many clicks", "Missing features", "Unreliable", "Let me explain"]
96
+
97
+ </using_askuserquestion>
98
+
99
+ <context_checklist>
100
+
101
+ Use this as a **background checklist**, not a conversation structure. Check these mentally as you go. If gaps remain, weave questions naturally.
102
+
103
+ - [ ] What they're building (concrete enough to explain to a stranger)
104
+ - [ ] Why it needs to exist (the problem or desire driving it)
105
+ - [ ] Who it's for (even if just themselves)
106
+ - [ ] What "done" looks like (observable outcomes)
107
+
108
+ Four things. If they volunteer more, capture it.
109
+
110
+ </context_checklist>
111
+
112
+ <decision_gate>
113
+
114
+ When you could write a clear PROJECT.md, offer to proceed:
115
+
116
+ - header: "Ready?"
117
+ - question: "I think I understand what you're after. Ready to create PROJECT.md?"
118
+ - options:
119
+ - "Create PROJECT.md" — Let's move forward
120
+ - "Keep exploring" — I want to share more / ask me more
121
+
122
+ If "Keep exploring" — ask what they want to add or identify gaps and probe naturally.
123
+
124
+ Loop until "Create PROJECT.md" selected.
125
+
126
+ </decision_gate>
127
+
128
+ <anti_patterns>
129
+
130
+ - **Checklist walking** — Going through domains regardless of what they said
131
+ - **Canned questions** — "What's your core value?" "What's out of scope?" regardless of context
132
+ - **Corporate speak** — "What are your success criteria?" "Who are your stakeholders?"
133
+ - **Interrogation** — Firing questions without building on answers
134
+ - **Rushing** — Minimizing questions to get to "the work"
135
+ - **Shallow acceptance** — Taking vague answers without probing
136
+ - **Premature constraints** — Asking about tech stack before understanding the idea
137
+ - **User skills** — NEVER ask about user's technical experience. Claude builds.
138
+
139
+ </anti_patterns>
140
+
141
+ </questioning_guide>