planning-workflow

Name: planning-workflow
Author: Dicklesworthstone

69stars13forks

How to install planning-workflow?

dicklesworthstone/agent_flywheel_clawdbot_skills_and_integrations/planning-workflow

$npx -y skills add Dicklesworthstone/agent_flywheel_clawdbot_skills_and_integrations --skill planning-workflow

Installs into the current project.

Summary

Jeffrey Emanuel's comprehensive markdown planning methodology for software projects. The 85%+ time-on-planning approach that makes agentic coding work at scale. Includes exact prompts used.

Files · 1

View on GitHub

SKILL.md

1	# Planning Workflow — The Foundation of Agentic Development
2
3	> Core Philosophy: "Planning tokens are a lot fewer and cheaper than implementation tokens."
4	>
5	> The models are far smarter when reasoning about a detailed plan that fits within their context window. This is the key insight behind spending 80%+ of time on planning.
6
7	---
8
9	## Why Planning Matters
10
11	Before burning tokens with a big agent swarm:
12
13	- Measure twice, cut once — becomes "Check your plan N times, implement once"
14	- A very big, complex markdown plan is still shorter than a few substantive code files
15	- Front-loading human input in planning enables removing yourself from implementation
16	- The code will be written ridiculously quickly when you start enough agents with a solid plan
17
18	---
19
20	## The Planning Process (Overview)
21
22	```
23	┌──────────────────────────────────────────────────────────────┐
24	│ 1. INITIAL PLAN (GPT Pro / Opus 4.5 in web app) │
25	│ └─► Explain goals, intent, workflows, tech stack │
26	├──────────────────────────────────────────────────────────────┤
27	│ 2. ITERATIVE REFINEMENT (GPT Pro Extended Reasoning) │
28	│ └─► 4-5 rounds of revision until steady-state │
29	├──────────────────────────────────────────────────────────────┤
30	│ 3. MULTI-MODEL BLENDING (Optional but recommended) │
31	│ └─► Gemini3 Deep Think, Grok4 Heavy, Opus 4.5 │
32	│ └─► GPT Pro as final arbiter │
33	├──────────────────────────────────────────────────────────────┤
34	│ 4. CONVERT TO BEADS (Claude Code + Opus 4.5) │
35	│ └─► Self-contained tasks with dependency structure │
36	├──────────────────────────────────────────────────────────────┤
37	│ 5. POLISH BEADS (6+ rounds until steady-state) │
38	│ └─► Cross-model review, never oversimplify │
39	└──────────────────────────────────────────────────────────────┘
40	```
41
42	---
43
44	## Phase 1: Creating the Initial Plan
45
46	### Where to Write It
47
48	Use GPT Pro with Extended Reasoning in the web app. No other model can touch Pro on the web when dealing with input that fits its context window.
49
50	Alternative: Claude Opus 4.5 in the webapp is also good for initial plans.
51
52	### What to Include
53
54	1. Goals and Intent — What you're really trying to accomplish
55	2. Workflows — How the final software should work from the user's perspective
56	3. Tech Stack — Be specific (e.g., "TypeScript, Next.js 16, React 19, Tailwind, Supabase")
57	4. Architecture Decisions — High-level structure and patterns
58	5. The "Why" — The more the model understands your end goal, the better it performs
59
60	You don't even need to write the initial markdown plan yourself. You can write that with GPT Pro, just explaining what it is you want to make.
61
62	---
63
64	## Phase 2: Iterative Refinement
65
66	### THE EXACT PROMPT — Plan Review (GPT Pro Extended Reasoning)
67
68	Paste your entire markdown plan into GPT Pro with Extended Reasoning enabled and use this EXACT prompt:
69
70	```
71	Carefully review this entire plan for me and come up with your best revisions in terms of better architecture, new features, changed features, etc. to make it better, more robust/reliable, more performant, more compelling/useful, etc. For each proposed change, give me your detailed analysis and rationale/justification for why it would make the project better along with the git-diff style change versus the original plan shown below:
72
73	<PASTE YOUR EXISTING COMPLETE PLAN HERE>
74	```
75
76	### THE EXACT PROMPT — Integration (Claude Code)
77
78	After GPT Pro finishes (may take 20-30 minutes for complex plans), paste the output into Claude Code with this EXACT prompt:
79
80	```
81	OK, now integrate these revisions to the markdown plan in-place; use ultrathink and be meticulous. At the end, you can tell me which changes you wholeheartedly agree with, which you somewhat agree with, and which you disagree with:
82
83	```[Pasted text from GPT Pro]```
84	```
85
86	### Repeat Until Steady-State
87
88	- Start fresh ChatGPT conversations for each round
89	- After 4-5 rounds, suggestions become very incremental
90	- You'll see massive improvements from v2 to v3, continuing to the end
91	- This phase can take 2-3 hours for complex features — this is normal
92
93	---
94
95	## Phase 3: Multi-Model Blending (Advanced)
96
97	### Why Blend Models
98
99	Different models have different strengths. Blending gets "best of all worlds."
100
101	### The Process
102
103	1. Get competing plans from Gemini3 (Deep Think), Grok4 Heavy, and Opus 4.5
104	2. Use GPT Pro as final arbiter
105
106	### THE EXACT PROMPT — Multi-Model Blend
107
108	```
109	I asked 3 competing LLMs to do the exact same thing and they came up with pretty different plans which you can read below. I want you to REALLY carefully analyze their plans with an open mind and be intellectually honest about what they

People who liked this also viewed

6 picks

Type

planning-workflow

69stars13forks

View on GitHub

How to install planning-workflow?

dicklesworthstone/agent_flywheel_clawdbot_skills_and_integrations/planning-workflow

$npx -y skills add Dicklesworthstone/agent_flywheel_clawdbot_skills_and_integrations --skill planning-workflow

Installs into the current project.

Summary

Jeffrey Emanuel's comprehensive markdown planning methodology for software projects. The 85%+ time-on-planning approach that makes agentic coding work at scale. Includes exact prompts used.

Files · 1

View on GitHub

SKILL.md

1	# Planning Workflow — The Foundation of Agentic Development
2
3	> Core Philosophy: "Planning tokens are a lot fewer and cheaper than implementation tokens."
4	>
5	> The models are far smarter when reasoning about a detailed plan that fits within their context window. This is the key insight behind spending 80%+ of time on planning.
6
7	---
8
9	## Why Planning Matters
10
11	Before burning tokens with a big agent swarm:
12
13	- Measure twice, cut once — becomes "Check your plan N times, implement once"
14	- A very big, complex markdown plan is still shorter than a few substantive code files
15	- Front-loading human input in planning enables removing yourself from implementation
16	- The code will be written ridiculously quickly when you start enough agents with a solid plan
17
18	---
19
20	## The Planning Process (Overview)
21
22	```
23	┌──────────────────────────────────────────────────────────────┐
24	│ 1. INITIAL PLAN (GPT Pro / Opus 4.5 in web app) │
25	│ └─► Explain goals, intent, workflows, tech stack │
26	├──────────────────────────────────────────────────────────────┤
27	│ 2. ITERATIVE REFINEMENT (GPT Pro Extended Reasoning) │
28	│ └─► 4-5 rounds of revision until steady-state │
29	├──────────────────────────────────────────────────────────────┤
30	│ 3. MULTI-MODEL BLENDING (Optional but recommended) │
31	│ └─► Gemini3 Deep Think, Grok4 Heavy, Opus 4.5 │
32	│ └─► GPT Pro as final arbiter │
33	├──────────────────────────────────────────────────────────────┤
34	│ 4. CONVERT TO BEADS (Claude Code + Opus 4.5) │
35	│ └─► Self-contained tasks with dependency structure │
36	├──────────────────────────────────────────────────────────────┤
37	│ 5. POLISH BEADS (6+ rounds until steady-state) │
38	│ └─► Cross-model review, never oversimplify │
39	└──────────────────────────────────────────────────────────────┘
40	```
41
42	---
43
44	## Phase 1: Creating the Initial Plan
45
46	### Where to Write It
47
48	Use GPT Pro with Extended Reasoning in the web app. No other model can touch Pro on the web when dealing with input that fits its context window.
49
50	Alternative: Claude Opus 4.5 in the webapp is also good for initial plans.
51
52	### What to Include
53
54	1. Goals and Intent — What you're really trying to accomplish
55	2. Workflows — How the final software should work from the user's perspective
56	3. Tech Stack — Be specific (e.g., "TypeScript, Next.js 16, React 19, Tailwind, Supabase")
57	4. Architecture Decisions — High-level structure and patterns
58	5. The "Why" — The more the model understands your end goal, the better it performs
59
60	You don't even need to write the initial markdown plan yourself. You can write that with GPT Pro, just explaining what it is you want to make.
61
62	---
63
64	## Phase 2: Iterative Refinement
65
66	### THE EXACT PROMPT — Plan Review (GPT Pro Extended Reasoning)
67
68	Paste your entire markdown plan into GPT Pro with Extended Reasoning enabled and use this EXACT prompt:
69
70	```
71	Carefully review this entire plan for me and come up with your best revisions in terms of better architecture, new features, changed features, etc. to make it better, more robust/reliable, more performant, more compelling/useful, etc. For each proposed change, give me your detailed analysis and rationale/justification for why it would make the project better along with the git-diff style change versus the original plan shown below:
72
73	<PASTE YOUR EXISTING COMPLETE PLAN HERE>
74	```
75
76	### THE EXACT PROMPT — Integration (Claude Code)
77
78	After GPT Pro finishes (may take 20-30 minutes for complex plans), paste the output into Claude Code with this EXACT prompt:
79
80	```
81	OK, now integrate these revisions to the markdown plan in-place; use ultrathink and be meticulous. At the end, you can tell me which changes you wholeheartedly agree with, which you somewhat agree with, and which you disagree with:
82
83	```[Pasted text from GPT Pro]```
84	```
85
86	### Repeat Until Steady-State
87
88	- Start fresh ChatGPT conversations for each round
89	- After 4-5 rounds, suggestions become very incremental
90	- You'll see massive improvements from v2 to v3, continuing to the end
91	- This phase can take 2-3 hours for complex features — this is normal
92
93	---
94
95	## Phase 3: Multi-Model Blending (Advanced)
96
97	### Why Blend Models
98
99	Different models have different strengths. Blending gets "best of all worlds."
100
101	### The Process
102
103	1. Get competing plans from Gemini3 (Deep Think), Grok4 Heavy, and Opus 4.5
104	2. Use GPT Pro as final arbiter
105
106	### THE EXACT PROMPT — Multi-Model Blend
107
108	```
109	I asked 3 competing LLMs to do the exact same thing and they came up with pretty different plans which you can read below. I want you to REALLY carefully analyze their plans with an open mind and be intellectually honest about what they

People who liked this also viewed

6 picks

Type