deep-research

Name: deep-research
Author: Undertone0809

247stars30forks

How to install deep-research?

undertone0809/rudder/deep-research

$npx -y skills add Undertone0809/rudder --skill deep-research

Installs into the current project.

Summary

Conducts enterprise-grade research with multi-source synthesis, citation tracking, and verification. Produces citation-backed reports through a structured pipeline with source credibility scoring, and grounds abstract findings in concrete examples, cases, counterexamples, or mini

Files · 1

View on GitHub

SKILL.md

1	# Deep Research
2
3	## Core Purpose
4
5	Deliver citation-backed, verified research reports through a structured pipeline with source credibility scoring, evidence persistence, and progressive context management.
6
7	Autonomy Principle: Operate independently. Infer assumptions from context. Only stop for critical errors or incomprehensible queries.
8
9	Accessibility Principle: Make dense analysis easier to grasp. When a finding is abstract, architectural, process-heavy, or strategically subtle, ground it in a concrete example, mini-case, counterexample, implementation scenario, or failure mode. Prefer real source-grounded cases; if you use an illustrative hypothetical, label it clearly as hypothetical and do not cite it as if it were sourced fact.
10
11	---
12
13	## Decision Tree
14
15	```
16	Request Analysis
17	+-- Simple lookup? --> STOP: Use WebSearch
18	+-- Debugging? --> STOP: Use standard tools
19	+-- Complex analysis needed? --> CONTINUE
20
21	Mode Selection
22	+-- Initial exploration --> quick (3 phases, 2-5 min)
23	+-- Standard research --> standard (6 phases, 5-10 min) [DEFAULT]
24	+-- Critical decision --> deep (8 phases, 10-20 min)
25	+-- Comprehensive review --> ultradeep (8+ phases, 20-45 min)
26	```
27
28	Default assumptions: Technical query = technical audience. Comparison = balanced perspective. Trend = recent 1-2 years.
29
30	Default readability rule: For Standard mode and above, try to include at least one concrete case/example for each major finding unless the topic is purely numeric, the source material provides no credible case material, or confidentiality/safety considerations make cases inappropriate.
31
32	---
33
34	## Workflow Overview
35
36	\| Phase \| Name \| Quick \| Standard \| Deep \| UltraDeep \|
37	\|-------\|------\|-------\|----------\|------\|-----------\|
38	\| 1 \| SCOPE \| Y \| Y \| Y \| Y \|
39	\| 2 \| PLAN \| - \| Y \| Y \| Y \|
40	\| 3 \| RETRIEVE \| Y \| Y \| Y \| Y \|
41	\| 4 \| TRIANGULATE \| - \| Y \| Y \| Y \|
42	\| 4.5 \| OUTLINE REFINEMENT \| - \| Y \| Y \| Y \|
43	\| 5 \| SYNTHESIZE \| - \| Y \| Y \| Y \|
44	\| 6 \| CRITIQUE \| - \| - \| Y \| Y \|
45	\| 7 \| REFINE \| - \| - \| Y \| Y \|
46	\| 8 \| PACKAGE \| Y \| Y \| Y \| Y \|
47
48	---
49
50	## Execution
51
52	On invocation, load relevant reference files:
53
54	1. Phase 1-7: Load [methodology.md](./reference/methodology.md) for detailed phase instructions
55	2. Phase 8 (Report): Load [report-assembly.md](./reference/report-assembly.md) for progressive generation
56	3. HTML/PDF output: Load [html-generation.md](./reference/html-generation.md)
57	4. Quality checks: Load [quality-gates.md](./reference/quality-gates.md)
58	5. Long reports (>18K words): Load [continuation.md](./reference/continuation.md)
59
60	Templates:
61	- Report structure: [report_template.md](./templates/report_template.md)
62	- HTML styling: [mckinsey_report_template.html](./templates/mckinsey_report_template.html)
63
64	Scripts:
65	- `python scripts/validate_report.py --report [path]`
66	- `python scripts/verify_citations.py --report [path]`
67	- `python scripts/md_to_html.py [markdown_path]`
68
69	---
70
71	## Output Contract
72
73	Required sections:
74	- Executive Summary (200-400 words)
75	- Introduction (scope, methodology, assumptions)
76	- Main Analysis (4-8 findings, 600-2,000 words each, cited, each grounded in concrete examples/cases when helpful)
77	- Synthesis & Insights (patterns, implications)
78	- Limitations & Caveats
79	- Recommendations
80	- Bibliography (COMPLETE - every citation, no placeholders)
81	- Methodology Appendix
82
83	Output files (all to `~/Documents/[Topic]_Research_[YYYYMMDD]/`):
84	- Markdown (primary source)
85	- HTML (McKinsey style, auto-opened)
86	- PDF (professional print, auto-opened)
87
88	Quality standards:
89	- 10+ sources, 3+ per major claim
90	- All claims cited immediately [N]
91	- No placeholders, no fabricated citations
92	- Prose-first (>=80%), bullets sparingly
93	- Use concrete examples, mini-cases, counterexamples, or applied scenarios to make major findings easier to understand
94	- Source-ground real cases when available; clearly label hypothetical illustrative examples instead of citing them as factual evidence
95
96	---
97
98	## When to Use / NOT Use
99
100	Use: Comprehensive analysis, technology comparisons, state-of-the-art reviews, multi-perspective investigation, market analysis.
101
102	Do NOT use: Simple lookups, debugging, 1-2 search answers, quick time-sensitive queries.

People who liked this also viewed

6 picks

Type

deep-research

247stars30forks

View on GitHub

How to install deep-research?

undertone0809/rudder/deep-research

$npx -y skills add Undertone0809/rudder --skill deep-research

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Deep Research
2
3	## Core Purpose
4
5	Deliver citation-backed, verified research reports through a structured pipeline with source credibility scoring, evidence persistence, and progressive context management.
6
7	Autonomy Principle: Operate independently. Infer assumptions from context. Only stop for critical errors or incomprehensible queries.
8
9	Accessibility Principle: Make dense analysis easier to grasp. When a finding is abstract, architectural, process-heavy, or strategically subtle, ground it in a concrete example, mini-case, counterexample, implementation scenario, or failure mode. Prefer real source-grounded cases; if you use an illustrative hypothetical, label it clearly as hypothetical and do not cite it as if it were sourced fact.
10
11	---
12
13	## Decision Tree
14
15	```
16	Request Analysis
17	+-- Simple lookup? --> STOP: Use WebSearch
18	+-- Debugging? --> STOP: Use standard tools
19	+-- Complex analysis needed? --> CONTINUE
20
21	Mode Selection
22	+-- Initial exploration --> quick (3 phases, 2-5 min)
23	+-- Standard research --> standard (6 phases, 5-10 min) [DEFAULT]
24	+-- Critical decision --> deep (8 phases, 10-20 min)
25	+-- Comprehensive review --> ultradeep (8+ phases, 20-45 min)
26	```
27
28	Default assumptions: Technical query = technical audience. Comparison = balanced perspective. Trend = recent 1-2 years.
29
30	Default readability rule: For Standard mode and above, try to include at least one concrete case/example for each major finding unless the topic is purely numeric, the source material provides no credible case material, or confidentiality/safety considerations make cases inappropriate.
31
32	---
33
34	## Workflow Overview
35
36	\| Phase \| Name \| Quick \| Standard \| Deep \| UltraDeep \|
37	\|-------\|------\|-------\|----------\|------\|-----------\|
38	\| 1 \| SCOPE \| Y \| Y \| Y \| Y \|
39	\| 2 \| PLAN \| - \| Y \| Y \| Y \|
40	\| 3 \| RETRIEVE \| Y \| Y \| Y \| Y \|
41	\| 4 \| TRIANGULATE \| - \| Y \| Y \| Y \|
42	\| 4.5 \| OUTLINE REFINEMENT \| - \| Y \| Y \| Y \|
43	\| 5 \| SYNTHESIZE \| - \| Y \| Y \| Y \|
44	\| 6 \| CRITIQUE \| - \| - \| Y \| Y \|
45	\| 7 \| REFINE \| - \| - \| Y \| Y \|
46	\| 8 \| PACKAGE \| Y \| Y \| Y \| Y \|
47
48	---
49
50	## Execution
51
52	On invocation, load relevant reference files:
53
54	1. Phase 1-7: Load [methodology.md](./reference/methodology.md) for detailed phase instructions
55	2. Phase 8 (Report): Load [report-assembly.md](./reference/report-assembly.md) for progressive generation
56	3. HTML/PDF output: Load [html-generation.md](./reference/html-generation.md)
57	4. Quality checks: Load [quality-gates.md](./reference/quality-gates.md)
58	5. Long reports (>18K words): Load [continuation.md](./reference/continuation.md)
59
60	Templates:
61	- Report structure: [report_template.md](./templates/report_template.md)
62	- HTML styling: [mckinsey_report_template.html](./templates/mckinsey_report_template.html)
63
64	Scripts:
65	- `python scripts/validate_report.py --report [path]`
66	- `python scripts/verify_citations.py --report [path]`
67	- `python scripts/md_to_html.py [markdown_path]`
68
69	---
70
71	## Output Contract
72
73	Required sections:
74	- Executive Summary (200-400 words)
75	- Introduction (scope, methodology, assumptions)
76	- Main Analysis (4-8 findings, 600-2,000 words each, cited, each grounded in concrete examples/cases when helpful)
77	- Synthesis & Insights (patterns, implications)
78	- Limitations & Caveats
79	- Recommendations
80	- Bibliography (COMPLETE - every citation, no placeholders)
81	- Methodology Appendix
82
83	Output files (all to `~/Documents/[Topic]_Research_[YYYYMMDD]/`):
84	- Markdown (primary source)
85	- HTML (McKinsey style, auto-opened)
86	- PDF (professional print, auto-opened)
87
88	Quality standards:
89	- 10+ sources, 3+ per major claim
90	- All claims cited immediately [N]
91	- No placeholders, no fabricated citations
92	- Prose-first (>=80%), bullets sparingly
93	- Use concrete examples, mini-cases, counterexamples, or applied scenarios to make major findings easier to understand
94	- Source-ground real cases when available; clearly label hypothetical illustrative examples instead of citing them as factual evidence
95
96	---
97
98	## When to Use / NOT Use
99
100	Use: Comprehensive analysis, technology comparisons, state-of-the-art reviews, multi-perspective investigation, market analysis.
101
102	Do NOT use: Simple lookups, debugging, 1-2 search answers, quick time-sensitive queries.

People who liked this also viewed

6 picks

Type