fusion-plan

Name: fusion-plan
Author: duolahypercho

457stars69forks

How to install fusion-plan?

duolahypercho/fusion-fable/fusion-plan

$npx -y skills add duolahypercho/fusion-fable --skill fusion-plan

Installs into the current project.

Summary

Produce a high-confidence, CONCISE implementation / architecture / strategy PLAN that follows the OMC plan system end to end. It first runs OMC's real requirements interview (auto-chains the omc-plan skill: one question at a time, explore-first, Analyst consult) to produce an i

Files · 1

View on GitHub

SKILL.md

1	# Fusion-Plan — OMC interview, then a 3-round iterative panel
2
3	This skill follows the OMC plan system: requirements are gathered by OMC's real interview FIRST, then
4	the fusion panel deepens the resulting plan, and the output flows back into OMC's review → execution
5	pipeline. The fusion panel replaces only the plan-thinking step; storage, format, the quality gate, and
6	execution all stay on OMC's rails.
7
8	**Hard rule (same as fusion): Opus 4.8 always judges and synthesizes. You are the judge — stay separate
9	from the panelists.** The Opus panelist is always a spawned subagent, never you.
10
11	## Step 0 — Requirements (pick the mode FIRST)
12
13	Decide interactive vs non-interactive before doing anything — this determines whether you may interview.
14
15	- Non-interactive when ANY of these is true: invoked with `--no-interview`; running autonomously while
16	executing a larger plan; running inside a spawned sub-agent; or you otherwise cannot reach a human who
17	will answer. This is the default during plan execution.
18	- Interactive only when a human is clearly present and able to answer right now — typically the user
19	typed `/fusion-plan ...` directly in the main session at the start of a planning session.
20
21	### Non-interactive mode (autonomous / sub-agent / `--no-interview`)
22
23	HARD RULE: do NOT interview. Do NOT call `AskUserQuestion`. Do NOT auto-chain the `omc-plan` interview.
24	There is no one to answer — blocking on a question would hang the run. Requirements were already gathered
25	upstream, so derive `SEED_PLAN` from the context that already exists, in this order:
26
27	1. The task/story doc if one exists (e.g. `docs/stories/STORY-<id>.md`).
28	2. The larger plan that flagged this task (e.g. the relevant `.omc/plans/*.md` section).
29	3. The task description / prompt args you were given.
30
31	If that context is genuinely insufficient to plan, do NOT guess and do NOT ask — stop and return a short
32	note listing exactly what's missing, so the human (or the orchestrator) resolves it. Then go to Step 1.
33
34	### Interactive mode (human present)
35
36	Run OMC's real requirements interview by invoking the OMC plan skill on the user's request verbatim:
37
38	```
39	Skill("oh-my-claudecode:omc-plan") # pass the user's request; do NOT pass --consensus
40	```
41
42	- Let OMC auto-detect interview vs direct mode (one question at a time via `AskUserQuestion`,
43	explore-first, Analyst consult). It writes an initial plan to `.omc/plans/<slug>.md`.
44	- Do NOT pass `--consensus` (redundant with the fusion panel) and do NOT let OMC hand off to
45	execution yet — stop at the plan.
46	- Capture the plan file path as `SEED_PLAN` (newest `.omc/plans/*.md` if not reported). It carries the
47	gathered requirements — they must survive into the final plan.
48	- If `omc-plan` is unavailable, fall back to a minimal inline interview (one question at a time).
49
50	## Step 1 — Preconditions
51
52	This skill targets the `opus4.8-gpt5.5` panel. Check codex is present:
53
54	```bash
55	command -v codex && codex --version
56	```
57
58	- If `codex` is installed → run the full Opus 4.8 + GPT-5.5 loop below.
59	- If `codex` is missing → tell the user, then fall back to two independent Opus 4.8 panelists per round
60	(`opus4.8-4.8`) rather than failing. Note the downgrade and how to enable GPT-5.5 (`npm i -g
61	@openai/codex` + `codex login`).
62
63	Read the shared references once (they live in the sibling skill):
64
65	```bash
66	cat ~/.claude/skills/fusion/references/panel.md
67	cat ~/.claude/skills/fusion/references/judge_rubric.md
68	```
69
70	Honor `panel.md`: every panelist gets the same input verbatim, no assigned "lenses" or personas.
71
72	## Step 2 — The loop, SEEDED from the OMC plan (round R = 1, 2, 3)
73
74	The panel does NOT draft blind — round 1 is already seeded by `SEED_PLAN`, so the interview's requirements
75	anchor everything.
76
77	### 2a. Build each panelist's prompt
78
79	- **Every round (1, 2, 3) — independent cri

People who liked this also viewed

6 picks

Type

fusion-plan

457stars69forks

View on GitHub

How to install fusion-plan?

duolahypercho/fusion-fable/fusion-plan

$npx -y skills add duolahypercho/fusion-fable --skill fusion-plan

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Fusion-Plan — OMC interview, then a 3-round iterative panel
2
3	This skill follows the OMC plan system: requirements are gathered by OMC's real interview FIRST, then
4	the fusion panel deepens the resulting plan, and the output flows back into OMC's review → execution
5	pipeline. The fusion panel replaces only the plan-thinking step; storage, format, the quality gate, and
6	execution all stay on OMC's rails.
7
8	**Hard rule (same as fusion): Opus 4.8 always judges and synthesizes. You are the judge — stay separate
9	from the panelists.** The Opus panelist is always a spawned subagent, never you.
10
11	## Step 0 — Requirements (pick the mode FIRST)
12
13	Decide interactive vs non-interactive before doing anything — this determines whether you may interview.
14
15	- Non-interactive when ANY of these is true: invoked with `--no-interview`; running autonomously while
16	executing a larger plan; running inside a spawned sub-agent; or you otherwise cannot reach a human who
17	will answer. This is the default during plan execution.
18	- Interactive only when a human is clearly present and able to answer right now — typically the user
19	typed `/fusion-plan ...` directly in the main session at the start of a planning session.
20
21	### Non-interactive mode (autonomous / sub-agent / `--no-interview`)
22
23	HARD RULE: do NOT interview. Do NOT call `AskUserQuestion`. Do NOT auto-chain the `omc-plan` interview.
24	There is no one to answer — blocking on a question would hang the run. Requirements were already gathered
25	upstream, so derive `SEED_PLAN` from the context that already exists, in this order:
26
27	1. The task/story doc if one exists (e.g. `docs/stories/STORY-<id>.md`).
28	2. The larger plan that flagged this task (e.g. the relevant `.omc/plans/*.md` section).
29	3. The task description / prompt args you were given.
30
31	If that context is genuinely insufficient to plan, do NOT guess and do NOT ask — stop and return a short
32	note listing exactly what's missing, so the human (or the orchestrator) resolves it. Then go to Step 1.
33
34	### Interactive mode (human present)
35
36	Run OMC's real requirements interview by invoking the OMC plan skill on the user's request verbatim:
37
38	```
39	Skill("oh-my-claudecode:omc-plan") # pass the user's request; do NOT pass --consensus
40	```
41
42	- Let OMC auto-detect interview vs direct mode (one question at a time via `AskUserQuestion`,
43	explore-first, Analyst consult). It writes an initial plan to `.omc/plans/<slug>.md`.
44	- Do NOT pass `--consensus` (redundant with the fusion panel) and do NOT let OMC hand off to
45	execution yet — stop at the plan.
46	- Capture the plan file path as `SEED_PLAN` (newest `.omc/plans/*.md` if not reported). It carries the
47	gathered requirements — they must survive into the final plan.
48	- If `omc-plan` is unavailable, fall back to a minimal inline interview (one question at a time).
49
50	## Step 1 — Preconditions
51
52	This skill targets the `opus4.8-gpt5.5` panel. Check codex is present:
53
54	```bash
55	command -v codex && codex --version
56	```
57
58	- If `codex` is installed → run the full Opus 4.8 + GPT-5.5 loop below.
59	- If `codex` is missing → tell the user, then fall back to two independent Opus 4.8 panelists per round
60	(`opus4.8-4.8`) rather than failing. Note the downgrade and how to enable GPT-5.5 (`npm i -g
61	@openai/codex` + `codex login`).
62
63	Read the shared references once (they live in the sibling skill):
64
65	```bash
66	cat ~/.claude/skills/fusion/references/panel.md
67	cat ~/.claude/skills/fusion/references/judge_rubric.md
68	```
69
70	Honor `panel.md`: every panelist gets the same input verbatim, no assigned "lenses" or personas.
71
72	## Step 2 — The loop, SEEDED from the OMC plan (round R = 1, 2, 3)
73
74	The panel does NOT draft blind — round 1 is already seeded by `SEED_PLAN`, so the interview's requirements
75	anchor everything.
76
77	### 2a. Build each panelist's prompt
78
79	- **Every round (1, 2, 3) — independent cri

People who liked this also viewed

6 picks

Type