create-agentic-starter 0.1.1 → 0.1.2
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/dist/index.js +190 -118
- package/package.json +1 -1
package/dist/index.js
CHANGED
|
@@ -1,179 +1,251 @@
|
|
|
1
1
|
#!/usr/bin/env node
|
|
2
|
-
import{mkdir as a,rm as
|
|
2
|
+
import{mkdir as a,rm as s,writeFile as i}from"fs/promises";import o from"path";var c=[".agentic","AGENTS.md"],m=()=>({"AGENTS.md":`# Agentic Starter Guide
|
|
3
3
|
|
|
4
|
-
This
|
|
4
|
+
This scaffold creates a reusable AI working area under \`.agentic/\` for project kick-off and pre-delivery documentation.
|
|
5
5
|
|
|
6
|
-
##
|
|
6
|
+
## What This Is For
|
|
7
7
|
|
|
8
|
-
|
|
9
|
-
2. Open your AI tool in this folder.
|
|
10
|
-
3. Start a new chat with \`@.agentic/init.md\`.
|
|
8
|
+
Use this scaffold when you want an AI tool to:
|
|
11
9
|
|
|
12
|
-
|
|
10
|
+
- understand a project from files, notes, screenshots, and links
|
|
11
|
+
- maintain structured project memory
|
|
12
|
+
- generate markdown deliverables such as BRD, FRD, estimate, and proposal
|
|
13
13
|
|
|
14
|
-
|
|
14
|
+
## Folder Purpose
|
|
15
15
|
|
|
16
|
-
|
|
17
|
-
|
|
18
|
-
|
|
19
|
-
|
|
20
|
-
|
|
21
|
-
|
|
22
|
-
node_modules
|
|
23
|
-
dist
|
|
24
|
-
.agentic/workspace/analysis
|
|
25
|
-
.agentic/workspace/documents
|
|
26
|
-
`,"AGENTS.md":`# Agent Instructions
|
|
16
|
+
- \`.agentic/init.md\`: first prompt for every new AI session
|
|
17
|
+
- \`.agentic/context.md\`: shared rules and folder map
|
|
18
|
+
- \`.agentic/commands/\`: task-specific prompt files you reference in your AI tool
|
|
19
|
+
- \`.agentic/workspace/project/\`: raw project inputs such as screenshots, notes, requirement docs, and reference links
|
|
20
|
+
- \`.agentic/workspace/memory/\`: evolving AI-maintained project understanding
|
|
21
|
+
- \`.agentic/workspace/documents/\`: polished markdown deliverables
|
|
27
22
|
|
|
28
|
-
|
|
23
|
+
## How To Use It
|
|
29
24
|
|
|
30
|
-
|
|
25
|
+
1. Open your AI tool in this project.
|
|
26
|
+
2. Start every new session with \`@.agentic/init.md\`.
|
|
27
|
+
3. Follow the command sequence below.
|
|
31
28
|
|
|
32
|
-
|
|
33
|
-
|
|
34
|
-
|
|
35
|
-
|
|
36
|
-
|
|
37
|
-
|
|
38
|
-
|
|
29
|
+
## V1 Command Sequence
|
|
30
|
+
|
|
31
|
+
1. \`@.agentic/init.md\`
|
|
32
|
+
2. \`@.agentic/commands/project-requirements.md\`
|
|
33
|
+
3. \`@.agentic/commands/architecture.md\`
|
|
34
|
+
4. \`@.agentic/commands/create-brd.md\`
|
|
35
|
+
5. \`@.agentic/commands/create-frd.md\`
|
|
36
|
+
6. \`@.agentic/commands/create-estimate.md\`
|
|
37
|
+
7. \`@.agentic/commands/create-proposal.md\`
|
|
38
|
+
|
|
39
|
+
If your tool does not support \`@file\` references, paste the contents of the prompt file into a new chat manually.
|
|
39
40
|
`,".agentic/init.md":`# Init
|
|
40
41
|
|
|
41
|
-
|
|
42
|
+
Read \`@.agentic/context.md\` first.
|
|
42
43
|
|
|
43
|
-
|
|
44
|
+
You are starting a fresh AI session inside this project. Your job is to build lightweight orientation without over-reading the repository.
|
|
44
45
|
|
|
45
|
-
|
|
46
|
-
2. Scan the full \`.agentic/workspace/\` directory and summarize what already exists.
|
|
47
|
-
3. Inspect \`.agentic/workspace/requirements/\` for screenshots, requirement notes, references, or client material.
|
|
48
|
-
4. If \`.agentic/workspace/analysis/project-overview.md\` already exists, review and improve it instead of replacing useful content.
|
|
49
|
-
5. If \`.agentic/workspace/analysis/architecture.md\` already exists, review and improve it instead of replacing useful content.
|
|
50
|
-
6. Ask the user only for information that is truly missing and blocks accurate analysis.
|
|
51
|
-
7. Create or update \`.agentic/workspace/analysis/project-overview.md\` with:
|
|
52
|
-
- project summary
|
|
53
|
-
- goals
|
|
54
|
-
- users or stakeholders
|
|
55
|
-
- major modules
|
|
56
|
-
- assumptions
|
|
57
|
-
- risks
|
|
58
|
-
- open questions
|
|
59
|
-
8. Create or update \`.agentic/workspace/analysis/architecture.md\` with an initial architecture direction if enough information exists.
|
|
60
|
-
9. End with a concise summary of what you found, what you created or updated, and tell the user:
|
|
61
|
-
|
|
62
|
-
\`Next type @.agentic/instructions/01-project-understanding.md\`
|
|
63
|
-
`,".agentic/context.md":"# Context Map\n\nThis repository is scaffolded for an AI-assisted delivery workflow.\n\n## Paths\n\n- `.agentic/init.md`: first prompt to run in a new chat.\n- `.agentic/context.md`: shared map every step should read before doing work.\n- `.agentic/instructions/`: ordered step prompts.\n- `.agentic/workspace/requirements/`: source inputs from client, discovery, screenshots, notes, and references.\n- `.agentic/workspace/requirements/figma-screenshots/`: raw Figma screenshots dropped by the team.\n- `.agentic/workspace/analysis/`: generated markdown outputs such as project understanding, architecture, BRD drafts, FRD drafts, and estimation notes.\n- `.agentic/workspace/documents/`: exported deliverables such as `.docx`, `.xlsx`, or PDFs.\n- `.agentic/workspace/code/`: implementation area for frontend, backend, shared packages, and related codebases.\n- `.agentic/workspace/scripts/`: helper generators such as BRD or estimation export scripts.\n\n## Working rules\n\n1. Always read this file before executing an instruction step.\n2. Reuse and improve existing files when present instead of blindly overwriting good work.\n3. Save outputs only in the folders described above.\n4. At the end of every step, tell the user the exact next prompt to run.\n",".agentic/instructions/01-project-understanding.md":`# Step 01: Project Understanding
|
|
46
|
+
## Read Scope
|
|
64
47
|
|
|
65
|
-
Read
|
|
48
|
+
1. Read only root-level files and folders in the current project.
|
|
49
|
+
2. Do not recursively inspect the full repository.
|
|
50
|
+
3. Treat \`.agentic/\` as the exception:
|
|
51
|
+
- inspect \`.agentic/workspace/project/\`
|
|
52
|
+
- inspect \`.agentic/workspace/memory/\`
|
|
53
|
+
- inspect \`.agentic/workspace/documents/\`
|
|
54
|
+
4. Use existing workspace context if it is already present.
|
|
66
55
|
|
|
67
|
-
|
|
68
|
-
- \`@.agentic/workspace/analysis/project-overview.md\` if it exists
|
|
69
|
-
- everything relevant inside \`@.agentic/workspace/requirements/\`
|
|
56
|
+
## What To Do
|
|
70
57
|
|
|
71
|
-
|
|
58
|
+
1. Summarize the root-level project structure.
|
|
59
|
+
2. Summarize what already exists in \`.agentic/workspace/project/\`, \`.agentic/workspace/memory/\`, and \`.agentic/workspace/documents/\`.
|
|
60
|
+
3. Build a lightweight understanding from:
|
|
61
|
+
- root-level structure
|
|
62
|
+
- existing workspace files
|
|
63
|
+
- the user's message in this chat
|
|
64
|
+
4. If the user mentions screenshots, notes, requirement docs, links, or file paths, tell them how those should be placed or referenced under \`.agentic/workspace/project/\`.
|
|
65
|
+
5. If project materials are missing, say what would be useful to add, but continue with what you can understand already.
|
|
66
|
+
6. Ask only blocking clarification questions.
|
|
67
|
+
7. End with:
|
|
72
68
|
|
|
73
|
-
|
|
74
|
-
|
|
75
|
-
3. Update or create \`.agentic/workspace/analysis/project-overview.md\` as a polished project-understanding document.
|
|
76
|
-
4. Keep the document practical and client-ready.
|
|
77
|
-
5. If information is missing, ask only the minimum blocking questions.
|
|
69
|
+
\`Next type @.agentic/commands/project-requirements.md\`
|
|
70
|
+
`,".agentic/context.md":`# Context
|
|
78
71
|
|
|
79
|
-
|
|
72
|
+
This repository uses a prompt-based project kick-off system.
|
|
80
73
|
|
|
81
|
-
|
|
74
|
+
## Folder Map
|
|
82
75
|
|
|
83
|
-
|
|
76
|
+
- \`.agentic/init.md\`: session bootstrap prompt
|
|
77
|
+
- \`.agentic/context.md\`: shared rules and folder semantics
|
|
78
|
+
- \`.agentic/commands/\`: prompt files for distinct tasks
|
|
79
|
+
- \`.agentic/workspace/project/\`: raw project materials
|
|
80
|
+
- \`.agentic/workspace/memory/\`: evolving AI-maintained understanding
|
|
81
|
+
- \`.agentic/workspace/documents/\`: polished markdown outputs
|
|
84
82
|
|
|
85
|
-
|
|
86
|
-
`,".agentic/instructions/02-architecture.md":`# Step 02: Architecture
|
|
83
|
+
## Workspace Semantics
|
|
87
84
|
|
|
88
|
-
|
|
85
|
+
### Project
|
|
89
86
|
|
|
90
|
-
|
|
91
|
-
- \`@.agentic/workspace/analysis/project-overview.md\`
|
|
92
|
-
- any relevant requirement material in \`@.agentic/workspace/requirements/\`
|
|
87
|
+
Use \`.agentic/workspace/project/\` for:
|
|
93
88
|
|
|
94
|
-
|
|
89
|
+
- Figma screenshots
|
|
90
|
+
- requirement notes
|
|
91
|
+
- requirement documents
|
|
92
|
+
- links captured into markdown files
|
|
93
|
+
- user-supplied project references
|
|
95
94
|
|
|
96
|
-
|
|
97
|
-
2. Cover frontend, backend, data, integrations, authentication, deployment, observability, and security where relevant.
|
|
98
|
-
3. Include a mermaid diagram if the AI tool supports it.
|
|
99
|
-
4. Update or create \`.agentic/workspace/analysis/architecture.md\`.
|
|
100
|
-
5. Keep it implementation-oriented and aligned with the project scope.
|
|
95
|
+
### Memory
|
|
101
96
|
|
|
102
|
-
|
|
97
|
+
Use \`.agentic/workspace/memory/\` for:
|
|
103
98
|
|
|
104
|
-
-
|
|
99
|
+
- \`project-overview.md\`
|
|
100
|
+
- \`requirements.md\`
|
|
101
|
+
- \`open-questions.md\`
|
|
102
|
+
- \`architecture.md\`
|
|
103
|
+
- assumptions and structured working memory
|
|
105
104
|
|
|
106
|
-
|
|
105
|
+
### Documents
|
|
106
|
+
|
|
107
|
+
Use \`.agentic/workspace/documents/\` for polished outputs:
|
|
107
108
|
|
|
108
|
-
\`
|
|
109
|
-
|
|
109
|
+
- \`brd.md\`
|
|
110
|
+
- \`frd.md\`
|
|
111
|
+
- \`estimate.md\`
|
|
112
|
+
- \`proposal.md\`
|
|
113
|
+
|
|
114
|
+
## Working Rules
|
|
115
|
+
|
|
116
|
+
1. Reuse and improve existing workspace files when they already exist.
|
|
117
|
+
2. Do not scan the whole repository unless the user explicitly asks.
|
|
118
|
+
3. Ask only the minimum blocking questions required to continue.
|
|
119
|
+
4. Always end by telling the user the exact next command to run.
|
|
120
|
+
`,".agentic/commands/project-requirements.md":`# Project Requirements
|
|
110
121
|
|
|
111
122
|
Read:
|
|
112
123
|
|
|
113
124
|
- \`@.agentic/context.md\`
|
|
114
|
-
-
|
|
115
|
-
-
|
|
116
|
-
-
|
|
125
|
+
- relevant existing files inside \`.agentic/workspace/project/\`
|
|
126
|
+
- \`.agentic/workspace/memory/project-overview.md\` if it exists
|
|
127
|
+
- \`.agentic/workspace/memory/requirements.md\` if it exists
|
|
128
|
+
- \`.agentic/workspace/memory/open-questions.md\` if it exists
|
|
129
|
+
|
|
130
|
+
## Your job
|
|
131
|
+
|
|
132
|
+
1. Collect the project explanation, goals, users, scope, screenshots, references, and known deliverables.
|
|
133
|
+
2. Accept either:
|
|
134
|
+
- existing file paths already shared by the user, or
|
|
135
|
+
- fresh explanation typed in chat
|
|
136
|
+
3. Tell the user where each input belongs under \`.agentic/workspace/project/\`.
|
|
137
|
+
4. Create or update:
|
|
138
|
+
- \`.agentic/workspace/memory/project-overview.md\`
|
|
139
|
+
- \`.agentic/workspace/memory/requirements.md\`
|
|
140
|
+
- \`.agentic/workspace/memory/open-questions.md\`
|
|
141
|
+
5. Capture assumptions separately from confirmed facts.
|
|
142
|
+
6. Ask only for missing details that block useful requirement understanding.
|
|
117
143
|
|
|
118
|
-
|
|
144
|
+
When done, tell the user:
|
|
119
145
|
|
|
120
|
-
|
|
121
|
-
|
|
122
|
-
3. Save the markdown draft to \`.agentic/workspace/analysis/brd.md\`.
|
|
123
|
-
4. If a generator script exists in \`.agentic/workspace/scripts/\`, mention whether it can be used to export a \`.docx\` into \`.agentic/workspace/documents/\`.
|
|
146
|
+
\`Next type @.agentic/commands/architecture.md\`
|
|
147
|
+
`,".agentic/commands/architecture.md":`# Architecture
|
|
124
148
|
|
|
125
|
-
|
|
149
|
+
Read:
|
|
126
150
|
|
|
127
|
-
-
|
|
151
|
+
- \`@.agentic/context.md\`
|
|
152
|
+
- \`.agentic/workspace/memory/project-overview.md\`
|
|
153
|
+
- \`.agentic/workspace/memory/requirements.md\`
|
|
154
|
+
- relevant files under \`.agentic/workspace/project/\`
|
|
155
|
+
- \`.agentic/workspace/memory/open-questions.md\` if it exists
|
|
156
|
+
|
|
157
|
+
## Your job
|
|
158
|
+
|
|
159
|
+
1. Produce or update \`.agentic/workspace/memory/architecture.md\`.
|
|
160
|
+
2. Explain the recommended system architecture based on the known requirements.
|
|
161
|
+
3. Cover frontend, backend, data, integrations, authentication, deployment, and key risks when relevant.
|
|
162
|
+
4. Add a diagram description or mermaid diagram when supported.
|
|
163
|
+
5. Mention draw.io MCP only as an optional tool if available in the user's environment.
|
|
164
|
+
6. Ask only the minimum blocking questions.
|
|
128
165
|
|
|
129
166
|
When done, tell the user:
|
|
130
167
|
|
|
131
|
-
\`Next type @.agentic/
|
|
132
|
-
`,".agentic/
|
|
168
|
+
\`Next type @.agentic/commands/create-brd.md\`
|
|
169
|
+
`,".agentic/commands/create-brd.md":`# Create BRD
|
|
133
170
|
|
|
134
171
|
Read:
|
|
135
172
|
|
|
136
173
|
- \`@.agentic/context.md\`
|
|
137
|
-
-
|
|
138
|
-
-
|
|
139
|
-
-
|
|
140
|
-
-
|
|
174
|
+
- \`.agentic/workspace/memory/project-overview.md\`
|
|
175
|
+
- \`.agentic/workspace/memory/requirements.md\`
|
|
176
|
+
- \`.agentic/workspace/memory/architecture.md\`
|
|
177
|
+
- relevant files under \`.agentic/workspace/project/\`
|
|
178
|
+
|
|
179
|
+
## Your job
|
|
180
|
+
|
|
181
|
+
1. Create or update \`.agentic/workspace/documents/brd.md\`.
|
|
182
|
+
2. Produce a polished markdown Business Requirements Document.
|
|
183
|
+
3. Use confirmed facts first and clearly label assumptions where needed.
|
|
184
|
+
4. If critical information is missing, ask only the minimum blocking questions before drafting.
|
|
185
|
+
|
|
186
|
+
When done, tell the user:
|
|
141
187
|
|
|
142
|
-
|
|
188
|
+
\`Next type @.agentic/commands/create-frd.md\`
|
|
189
|
+
`,".agentic/commands/create-frd.md":`# Create FRD
|
|
143
190
|
|
|
144
|
-
|
|
145
|
-
2. Cover modules, user flows, functional requirements, validations, edge cases, roles, permissions, reporting needs, and non-functional considerations where relevant.
|
|
146
|
-
3. Save the markdown draft to \`.agentic/workspace/analysis/frd.md\`.
|
|
147
|
-
4. Keep the structure ready for handoff to product, design, and engineering teams.
|
|
191
|
+
Read:
|
|
148
192
|
|
|
149
|
-
|
|
193
|
+
- \`@.agentic/context.md\`
|
|
194
|
+
- \`.agentic/workspace/memory/project-overview.md\`
|
|
195
|
+
- \`.agentic/workspace/memory/requirements.md\`
|
|
196
|
+
- \`.agentic/workspace/memory/architecture.md\`
|
|
197
|
+
- \`.agentic/workspace/documents/brd.md\` if it exists
|
|
198
|
+
- relevant files under \`.agentic/workspace/project/\`
|
|
199
|
+
|
|
200
|
+
## Your job
|
|
150
201
|
|
|
151
|
-
|
|
202
|
+
1. Create or update \`.agentic/workspace/documents/frd.md\`.
|
|
203
|
+
2. Produce a polished markdown Functional Requirements Document.
|
|
204
|
+
3. Cover modules, flows, validations, roles, and edge cases where relevant.
|
|
205
|
+
4. Ask only the minimum blocking questions.
|
|
152
206
|
|
|
153
207
|
When done, tell the user:
|
|
154
208
|
|
|
155
|
-
\`Next type @.agentic/
|
|
156
|
-
`,".agentic/
|
|
209
|
+
\`Next type @.agentic/commands/create-estimate.md\`
|
|
210
|
+
`,".agentic/commands/create-estimate.md":`# Create Estimate
|
|
157
211
|
|
|
158
212
|
Read:
|
|
159
213
|
|
|
160
214
|
- \`@.agentic/context.md\`
|
|
161
|
-
-
|
|
162
|
-
-
|
|
163
|
-
-
|
|
164
|
-
-
|
|
215
|
+
- \`.agentic/workspace/memory/project-overview.md\`
|
|
216
|
+
- \`.agentic/workspace/memory/requirements.md\`
|
|
217
|
+
- \`.agentic/workspace/memory/architecture.md\`
|
|
218
|
+
- \`.agentic/workspace/documents/brd.md\` if it exists
|
|
219
|
+
- \`.agentic/workspace/documents/frd.md\` if it exists
|
|
165
220
|
|
|
166
|
-
Your job
|
|
221
|
+
## Your job
|
|
222
|
+
|
|
223
|
+
1. Create or update \`.agentic/workspace/documents/estimate.md\`.
|
|
224
|
+
2. Produce a markdown estimate with effort breakdown, assumptions, dependencies, and risks.
|
|
225
|
+
3. Keep the estimate aligned with the known scope and clearly state uncertainty.
|
|
226
|
+
4. Ask only the minimum blocking questions.
|
|
167
227
|
|
|
168
|
-
|
|
169
|
-
2. Break effort down by module, role, and phase where possible.
|
|
170
|
-
3. Include assumptions, dependencies, delivery risks, and scope sensitivity.
|
|
171
|
-
4. Save the markdown draft to \`.agentic/workspace/analysis/estimation.md\`.
|
|
172
|
-
5. If the workspace includes generator scripts, mention whether the estimate can be exported into \`.agentic/workspace/documents/\`.
|
|
228
|
+
When done, tell the user:
|
|
173
229
|
|
|
174
|
-
|
|
230
|
+
\`Next type @.agentic/commands/create-proposal.md\`
|
|
231
|
+
`,".agentic/commands/create-proposal.md":`# Create Proposal
|
|
232
|
+
|
|
233
|
+
Read:
|
|
234
|
+
|
|
235
|
+
- \`@.agentic/context.md\`
|
|
236
|
+
- \`.agentic/workspace/memory/project-overview.md\`
|
|
237
|
+
- \`.agentic/workspace/memory/requirements.md\`
|
|
238
|
+
- \`.agentic/workspace/memory/architecture.md\`
|
|
239
|
+
- \`.agentic/workspace/documents/brd.md\` if it exists
|
|
240
|
+
- \`.agentic/workspace/documents/frd.md\` if it exists
|
|
241
|
+
- \`.agentic/workspace/documents/estimate.md\` if it exists
|
|
242
|
+
|
|
243
|
+
## Your job
|
|
175
244
|
|
|
176
|
-
|
|
245
|
+
1. Create or update \`.agentic/workspace/documents/proposal.md\`.
|
|
246
|
+
2. Produce a polished markdown proposal using the known scope, architecture direction, and estimate.
|
|
247
|
+
3. Make gaps explicit instead of inventing details.
|
|
248
|
+
4. Ask only the minimum blocking questions.
|
|
177
249
|
|
|
178
|
-
When done, tell the user the workflow is complete and list the generated
|
|
179
|
-
`}),
|
|
250
|
+
When done, tell the user the v1 workflow is complete and list the generated files.
|
|
251
|
+
`}),d=[".agentic",".agentic/commands",".agentic/workspace",".agentic/workspace/project",".agentic/workspace/memory",".agentic/workspace/documents"];async function p(e){await Promise.all(c.map(t=>s(o.join(e,t),{recursive:!0,force:!0})))}async function l(e){await Promise.all(d.map(t=>a(o.join(e,t),{recursive:!0})))}async function u(e){let t=m();await Promise.all(Object.entries(t).map(([r,n])=>i(o.join(e,r),n,"utf8")))}async function g(){let e=process.cwd();await p(e),await l(e),await u(e),console.log(""),console.log("create-agentic-starter: scaffold created successfully."),console.log(`Project folder: ${e}`),console.log(""),console.log("Next steps:"),console.log("1. Open your AI tool in this folder."),console.log("2. Start a new chat with @.agentic/init.md"),console.log("")}g().catch(e=>{let t=e instanceof Error?e.message:String(e);console.error(`create-agentic-starter: failed to scaffold project. ${t}`),process.exitCode=1});
|