tree-of-thoughts

Name: tree-of-thoughts
Author: NeoLabHQ

853installs132forks

How to install tree-of-thoughts?

neolabhq/context-engineering-kit/tree-of-thoughts

$npx -y skills add NeoLabHQ/context-engineering-kit --skill tree-of-thoughts

Installs into the current project.

Summary

Execute tasks through systematic exploration, pruning, and expansion using Tree of Thoughts methodology with meta-judge evaluation specifications and multi-agent evaluation

Files · 1

View on GitHub

SKILL.md

1	# tree-of-thoughts
2
3	<task>
4	Execute complex reasoning tasks through systematic exploration of solution space, pruning unpromising branches, expanding viable approaches, and synthesizing the best solution.
5	</task>
6
7	<context>
8	This command implements the Tree of Thoughts (ToT) pattern for tasks requiring exploration of multiple solution paths before committing to full implementation. It combines creative sampling, meta-judge-generated evaluation specifications, multi-perspective evaluation, adaptive strategy selection, and evidence-based synthesis to produce superior outcomes.
9
10	Key benefits:
11
12	- Systematic exploration - Multiple agents explore different regions of the solution space
13	- Structured evaluation - Meta-judges produce tailored rubrics and criteria before judging
14	- Independent verification - Judges apply meta-judge specifications mechanically, reducing bias
15	- Adaptive strategy - Clear winners get polished, split decisions get synthesized, failures get redesigned
16	</context>
17
18	## Pattern: Tree of Thoughts (ToT)
19
20	This command implements an eight-phase systematic reasoning pattern with meta-judge evaluation and adaptive strategy selection:
21
22	```
23	Phase 1: Exploration (Propose Approaches)
24	┌─ Agent A → Proposals A1, A2 (with probabilities) ─┐
25	Task ───┼─ Agent B → Proposals B1, B2 (with probabilities) ─┼─┐
26	└─ Agent C → Proposals C1, C2 (with probabilities) ─┘ │
27	│
28	Phase 1.5: Pruning Meta-Judge (runs in parallel with Phase 1) │
29	Meta-Judge → Pruning Evaluation Specification YAML ───┤
30	│
31	Phase 2: Pruning (Vote for Best 3) │
32	┌─ Judge 1 → Votes + Rationale ─┐ │
33	├─ Judge 2 → Votes + Rationale ─┼─────────────────────┤
34	└─ Judge 3 → Votes + Rationale ─┘ │
35	│ │
36	├─→ Select Top 3 Proposals │
37	│ │
38	Phase 3: Expansion (Develop Full Solutions) │
39	┌─ Agent A → Solution A (from proposal X) ─┐ │
40	├─ Agent B → Solution B (from proposal Y) ─┼──────────┤
41	└─ Agent C → Solution C (from proposal Z) ─┘ │
42	│
43	Phase 3.5: Evaluation Meta-Judge (runs in parallel w/ Phase 3)│
44	Meta-Judge → Evaluation Specification YAML ───────────┤
45	│
46	Phase 4: Evaluation (Judge Full Solutions) │
47	┌─ Judge 1 → Report 1 ─┐ │
48	├─ Judge 2 → Report 2 ─┼──────────────────────────────┤
49	└─ Judge 3 → Report 3 ─┘ │
50	│
51	Phase 4.5: Adaptive Strategy Selection │
52	Analyze Consensus ────────────────────────────────────┤
53	├─ Clear Winner? → SELECT_AND_POLISH │
54	├─ All Flawed (<3.0)? → REDESIGN (Phase 3) │
55	└─ Split Decision? → FULL_SYNTHESIS │
56	│ │
57	Phase 5: Synthesis (Only if FULL_SYNTHESIS) │
58	Synthesizer ────────────────────┴──────────────────────┴─→ Final Solution
59	```
60
61	## Process
62
63	### Setup: Create Directory Structure
64
65	Before starting, ensure the directory structure exists:
66
67	```bash
68	mkdir -p .specs/research .specs/reports
69	```
70
71	Naming conventions:
72	- Proposals: `.specs/research/{solution-name}-{YYYY-MM-DD}.proposals.[a\|b\|c].md`
73	- Pruning: `.specs/research/{solution-name}-{YYYY-MM-DD}.pruning.[1\|2\|3].md`
74	- Selection: `.specs/research/{solution-name}-{YYYY-MM-DD}.selection.md`
75	- Evaluation: `.specs/reports/{solution-name}-{YYYY-MM-DD}.[1\|2\|3].md`
76
77	Where:
78	- `{solution-name}` - Derived from output path (e.g., `users-api` from output `specs/api/users.md`)
79	- `{YYYY-MM-DD}` - Current date
80
81	Note: Solutions remain in their specified output locations; only research and evaluation files go to `.specs/`
82
83	### Phase 1: Exploration (Propose Approaches)
84
85	Launch 3 independent agents in parallel (recommended: Sonnet for speed):
86
87	1. Each agent receives identical task description and context
88	2. Each agent generates 6 high-level approaches (not full implementations)
89	3. For each approach, agent provides:
90	- Approach description (2-3 paragraphs)
91	- Key design decisions and trade-offs
92	- Probability estimate (0.0-1.0)
93	- **Est

People who liked this also viewed

6 picks

Type

tree-of-thoughts

853installs132forks

View on GitHub

How to install tree-of-thoughts?

neolabhq/context-engineering-kit/tree-of-thoughts

$npx -y skills add NeoLabHQ/context-engineering-kit --skill tree-of-thoughts

Installs into the current project.

Summary

Execute tasks through systematic exploration, pruning, and expansion using Tree of Thoughts methodology with meta-judge evaluation specifications and multi-agent evaluation

Files · 1

View on GitHub

SKILL.md

1	# tree-of-thoughts
2
3	<task>
4	Execute complex reasoning tasks through systematic exploration of solution space, pruning unpromising branches, expanding viable approaches, and synthesizing the best solution.
5	</task>
6
7	<context>
8	This command implements the Tree of Thoughts (ToT) pattern for tasks requiring exploration of multiple solution paths before committing to full implementation. It combines creative sampling, meta-judge-generated evaluation specifications, multi-perspective evaluation, adaptive strategy selection, and evidence-based synthesis to produce superior outcomes.
9
10	Key benefits:
11
12	- Systematic exploration - Multiple agents explore different regions of the solution space
13	- Structured evaluation - Meta-judges produce tailored rubrics and criteria before judging
14	- Independent verification - Judges apply meta-judge specifications mechanically, reducing bias
15	- Adaptive strategy - Clear winners get polished, split decisions get synthesized, failures get redesigned
16	</context>
17
18	## Pattern: Tree of Thoughts (ToT)
19
20	This command implements an eight-phase systematic reasoning pattern with meta-judge evaluation and adaptive strategy selection:
21
22	```
23	Phase 1: Exploration (Propose Approaches)
24	┌─ Agent A → Proposals A1, A2 (with probabilities) ─┐
25	Task ───┼─ Agent B → Proposals B1, B2 (with probabilities) ─┼─┐
26	└─ Agent C → Proposals C1, C2 (with probabilities) ─┘ │
27	│
28	Phase 1.5: Pruning Meta-Judge (runs in parallel with Phase 1) │
29	Meta-Judge → Pruning Evaluation Specification YAML ───┤
30	│
31	Phase 2: Pruning (Vote for Best 3) │
32	┌─ Judge 1 → Votes + Rationale ─┐ │
33	├─ Judge 2 → Votes + Rationale ─┼─────────────────────┤
34	└─ Judge 3 → Votes + Rationale ─┘ │
35	│ │
36	├─→ Select Top 3 Proposals │
37	│ │
38	Phase 3: Expansion (Develop Full Solutions) │
39	┌─ Agent A → Solution A (from proposal X) ─┐ │
40	├─ Agent B → Solution B (from proposal Y) ─┼──────────┤
41	└─ Agent C → Solution C (from proposal Z) ─┘ │
42	│
43	Phase 3.5: Evaluation Meta-Judge (runs in parallel w/ Phase 3)│
44	Meta-Judge → Evaluation Specification YAML ───────────┤
45	│
46	Phase 4: Evaluation (Judge Full Solutions) │
47	┌─ Judge 1 → Report 1 ─┐ │
48	├─ Judge 2 → Report 2 ─┼──────────────────────────────┤
49	└─ Judge 3 → Report 3 ─┘ │
50	│
51	Phase 4.5: Adaptive Strategy Selection │
52	Analyze Consensus ────────────────────────────────────┤
53	├─ Clear Winner? → SELECT_AND_POLISH │
54	├─ All Flawed (<3.0)? → REDESIGN (Phase 3) │
55	└─ Split Decision? → FULL_SYNTHESIS │
56	│ │
57	Phase 5: Synthesis (Only if FULL_SYNTHESIS) │
58	Synthesizer ────────────────────┴──────────────────────┴─→ Final Solution
59	```
60
61	## Process
62
63	### Setup: Create Directory Structure
64
65	Before starting, ensure the directory structure exists:
66
67	```bash
68	mkdir -p .specs/research .specs/reports
69	```
70
71	Naming conventions:
72	- Proposals: `.specs/research/{solution-name}-{YYYY-MM-DD}.proposals.[a\|b\|c].md`
73	- Pruning: `.specs/research/{solution-name}-{YYYY-MM-DD}.pruning.[1\|2\|3].md`
74	- Selection: `.specs/research/{solution-name}-{YYYY-MM-DD}.selection.md`
75	- Evaluation: `.specs/reports/{solution-name}-{YYYY-MM-DD}.[1\|2\|3].md`
76
77	Where:
78	- `{solution-name}` - Derived from output path (e.g., `users-api` from output `specs/api/users.md`)
79	- `{YYYY-MM-DD}` - Current date
80
81	Note: Solutions remain in their specified output locations; only research and evaluation files go to `.specs/`
82
83	### Phase 1: Exploration (Propose Approaches)
84
85	Launch 3 independent agents in parallel (recommended: Sonnet for speed):
86
87	1. Each agent receives identical task description and context
88	2. Each agent generates 6 high-level approaches (not full implementations)
89	3. For each approach, agent provides:
90	- Approach description (2-3 paragraphs)
91	- Key design decisions and trade-offs
92	- Probability estimate (0.0-1.0)
93	- **Est

People who liked this also viewed

6 picks

Type