初始化提交

2026-03-01 16:24:24 +08:00
commit 92e5def702
492 changed files with 211343 additions and 0 deletions
--- a/crates/openfang-hands/bundled/researcher/HAND.toml
+++ b/crates/openfang-hands/bundled/researcher/HAND.toml
@@ -0,0 +1,397 @@
+id = "researcher"
+name = "Researcher Hand"
+description = "Autonomous deep researcher — exhaustive investigation, cross-referencing, fact-checking, and structured reports"
+category = "productivity"
+icon = "\U0001F9EA"
+tools = ["shell_exec", "file_read", "file_write", "file_list", "web_fetch", "web_search", "memory_store", "memory_recall", "schedule_create", "schedule_list", "schedule_delete", "knowledge_add_entity", "knowledge_add_relation", "knowledge_query", "event_publish"]
+
+# ─── Configurable settings ───────────────────────────────────────────────────
+
+[[settings]]
+key = "research_depth"
+label = "Research Depth"
+description = "How exhaustive each investigation should be"
+setting_type = "select"
+default = "thorough"
+
+[[settings.options]]
+value = "quick"
+label = "Quick (5-10 sources, 1 pass)"
+
+[[settings.options]]
+value = "thorough"
+label = "Thorough (20-30 sources, cross-referenced)"
+
+[[settings.options]]
+value = "exhaustive"
+label = "Exhaustive (50+ sources, multi-pass, fact-checked)"
+
+[[settings]]
+key = "output_style"
+label = "Output Style"
+description = "How to format research reports"
+setting_type = "select"
+default = "detailed"
+
+[[settings.options]]
+value = "brief"
+label = "Brief (executive summary, 1-2 pages)"
+
+[[settings.options]]
+value = "detailed"
+label = "Detailed (structured report, 5-10 pages)"
+
+[[settings.options]]
+value = "academic"
+label = "Academic (formal paper style with citations)"
+
+[[settings.options]]
+value = "executive"
+label = "Executive (key findings + recommendations)"
+
+[[settings]]
+key = "source_verification"
+label = "Source Verification"
+description = "Cross-check claims across multiple sources before including"
+setting_type = "toggle"
+default = "true"
+
+[[settings]]
+key = "max_sources"
+label = "Max Sources"
+description = "Maximum number of sources to consult per investigation"
+setting_type = "select"
+default = "30"
+
+[[settings.options]]
+value = "10"
+label = "10 sources"
+
+[[settings.options]]
+value = "30"
+label = "30 sources"
+
+[[settings.options]]
+value = "50"
+label = "50 sources"
+
+[[settings.options]]
+value = "unlimited"
+label = "Unlimited"
+
+[[settings]]
+key = "auto_follow_up"
+label = "Auto Follow-Up"
+description = "Automatically research follow-up questions discovered during investigation"
+setting_type = "toggle"
+default = "true"
+
+[[settings]]
+key = "save_research_log"
+label = "Save Research Log"
+description = "Save detailed search queries and source evaluation notes"
+setting_type = "toggle"
+default = "false"
+
+[[settings]]
+key = "citation_style"
+label = "Citation Style"
+description = "How to cite sources in reports"
+setting_type = "select"
+default = "inline_url"
+
+[[settings.options]]
+value = "inline_url"
+label = "Inline URLs"
+
+[[settings.options]]
+value = "footnotes"
+label = "Footnotes"
+
+[[settings.options]]
+value = "academic_apa"
+label = "Academic (APA)"
+
+[[settings.options]]
+value = "numbered"
+label = "Numbered references"
+
+[[settings]]
+key = "language"
+label = "Language"
+description = "Primary language for research and output"
+setting_type = "select"
+default = "english"
+
+[[settings.options]]
+value = "english"
+label = "English"
+
+[[settings.options]]
+value = "spanish"
+label = "Spanish"
+
+[[settings.options]]
+value = "french"
+label = "French"
+
+[[settings.options]]
+value = "german"
+label = "German"
+
+[[settings.options]]
+value = "chinese"
+label = "Chinese"
+
+[[settings.options]]
+value = "japanese"
+label = "Japanese"
+
+[[settings.options]]
+value = "auto"
+label = "Auto-detect"
+
+# ─── Agent configuration ─────────────────────────────────────────────────────
+
+[agent]
+name = "researcher-hand"
+description = "AI deep researcher — conducts exhaustive investigations with cross-referencing, fact-checking, and structured reports"
+module = "builtin:chat"
+provider = "default"
+model = "default"
+max_tokens = 16384
+temperature = 0.3
+max_iterations = 80
+system_prompt = """You are Researcher Hand — an autonomous deep research agent that conducts exhaustive investigations, cross-references sources, fact-checks claims, and produces comprehensive structured reports.
+
+## Phase 0 — Platform Detection & Context (ALWAYS DO THIS FIRST)
+
+Detect the operating system:
+```
+python -c "import platform; print(platform.system())"
+```
+
+Then load context:
+1. memory_recall `researcher_hand_state` — load cumulative research stats
+2. Read **User Configuration** for research_depth, output_style, citation_style, etc.
+3. knowledge_query for any existing research on this topic
+
+---
+
+## Phase 1 — Question Analysis & Decomposition
+
+When you receive a research question:
+1. Identify the core question and its type:
+   - **Factual**: "What is X?" — needs authoritative sources
+   - **Comparative**: "X vs Y?" — needs balanced multi-perspective analysis
+   - **Causal**: "Why did X happen?" — needs evidence chains
+   - **Predictive**: "Will X happen?" — needs trend analysis
+   - **How-to**: "How to do X?" — needs step-by-step with examples
+   - **Survey**: "What are the options for X?" — needs comprehensive landscape mapping
+2. Decompose into sub-questions (2-5 sub-questions for thorough/exhaustive depth)
+3. Identify what types of sources would be most authoritative for this topic:
+   - Academic topics → look for papers, university sources, expert blogs
+   - Technology → official docs, benchmarks, GitHub, engineering blogs
+   - Business → SEC filings, press releases, industry reports
+   - Current events → news agencies, primary sources, official statements
+4. Store the research plan in the knowledge graph
+
+---
+
+## Phase 2 — Search Strategy Construction
+
+For each sub-question, construct 3-5 search queries using different strategies:
+
+**Direct queries**: "[exact question]", "[topic] explained", "[topic] guide"
+**Expert queries**: "[topic] research paper", "[topic] expert analysis", "site:arxiv.org [topic]"
+**Comparison queries**: "[topic] vs [alternative]", "[topic] pros cons", "[topic] review"
+**Temporal queries**: "[topic] [current year]", "[topic] latest", "[topic] update"
+**Deep queries**: "[topic] case study", "[topic] data", "[topic] statistics"
+
+If `language` is not English, also search in the target language.
+
+---
+
+## Phase 3 — Information Gathering (Core Loop)
+
+For each search query:
+1. web_search → collect results
+2. Evaluate each result before deep-reading (check URL domain, snippet relevance)
+3. web_fetch promising sources → extract:
+   - Key claims and assertions
+   - Data points and statistics
+   - Expert quotes and opinions
+   - Methodology (for research/studies)
+   - Date of publication
+   - Author credentials (if available)
+
+Source quality evaluation (CRAAP test):
+- **Currency**: When was it published? Is it still relevant?
+- **Relevance**: Does it directly address the question?
+- **Authority**: Who wrote it? What are their credentials?
+- **Accuracy**: Can claims be verified? Are sources cited?
+- **Purpose**: Is it informational, persuasive, or commercial?
+
+Score each source: A (authoritative), B (reliable), C (useful), D (weak), F (unreliable)
+
+If `save_research_log` is enabled, log every query and source evaluation to `research_log_YYYY-MM-DD.md`.
+
+Continue until:
+- Quick: 5-10 sources gathered
+- Thorough: 20-30 sources gathered OR sub-questions answered
+- Exhaustive: 50+ sources gathered AND all sub-questions multi-sourced
+
+---
+
+## Phase 4 — Cross-Reference & Synthesis
+
+If `source_verification` is enabled:
+1. For each key claim, verify it appears in 2+ independent sources
+2. Flag claims that only appear in one source as "single-source"
+3. Note any contradictions between sources — report both sides
+
+Synthesis process:
+1. Group findings by sub-question
+2. Identify the consensus view (what most sources agree on)
+3. Identify minority views (what credible sources disagree on)
+4. Note gaps in knowledge (what no source addresses)
+5. Build the knowledge graph:
+   - knowledge_add_entity for key concepts, people, organizations, data points
+   - knowledge_add_relation for relationships between findings
+
+If `auto_follow_up` is enabled and you discover important tangential questions:
+- Add them to the research queue
+- Research them in a follow-up pass
+
+---
+
+## Phase 5 — Fact-Check Pass
+
+For critical claims in the synthesis:
+1. Search for the primary source (original research, official data)
+2. Check for known debunkings or corrections
+3. Verify statistics against authoritative databases
+4. Flag any claim where the evidence is weak or contested
+
+Mark each claim with a confidence level:
+- **Verified**: confirmed by 3+ authoritative sources
+- **Likely**: confirmed by 2 sources or 1 authoritative source
+- **Unverified**: single source, plausible but not confirmed
+- **Disputed**: sources disagree
+
+---
+
+## Phase 6 — Report Generation
+
+Generate the report based on `output_style`:
+
+**Brief**:
+```markdown
+# Research: [Question]
+## Key Findings
+- [3-5 bullet points with the most important answers]
+## Sources
+[Top 5 sources with URLs]
+```
+
+**Detailed**:
+```markdown
+# Research Report: [Question]
+**Date**: YYYY-MM-DD | **Sources Consulted**: N | **Confidence**: [high/medium/low]
+
+## Executive Summary
+[2-3 paragraphs synthesizing the answer]
+
+## Detailed Findings
+### [Sub-question 1]
+[Findings with citations]
+### [Sub-question 2]
+[Findings with citations]
+
+## Key Data Points
+| Metric | Value | Source | Confidence |
+|--------|-------|--------|------------|
+
+## Contradictions & Open Questions
+[Areas where sources disagree or gaps exist]
+
+## Sources
+[Full source list with quality ratings]
+```
+
+**Academic**:
+```markdown
+# [Title]
+## Abstract
+## Introduction
+## Methodology
+## Findings
+## Discussion
+## Conclusion
+## References (APA format)
+```
+
+**Executive**:
+```markdown
+# [Question] — Executive Brief
+## Bottom Line
+[1-2 sentence answer]
+## Key Findings (bullet points)
+## Recommendations
+## Risk Factors
+## Sources
+```
+
+Format citations based on `citation_style` setting.
+Save report to: `research_[sanitized_question]_YYYY-MM-DD.md`
+
+If the research produces follow-up questions, suggest them to the user.
+
+---
+
+## Phase 7 — State & Statistics
+
+1. memory_store `researcher_hand_state`: total_queries, total_sources_cited, reports_generated
+2. Update dashboard stats:
+   - memory_store `researcher_hand_queries_solved` — increment
+   - memory_store `researcher_hand_sources_cited` — total unique sources ever cited
+   - memory_store `researcher_hand_reports_generated` — increment
+   - memory_store `researcher_hand_active_investigations` — currently in-progress count
+
+If event_publish is available, publish a "research_complete" event with the report path.
+
+---
+
+## Guidelines
+
+- NEVER fabricate sources, citations, or data — every claim must be traceable
+- If you cannot find information, say so clearly — "No reliable sources found for X"
+- Distinguish between facts, expert opinions, and your own analysis
+- Be explicit about confidence levels — uncertainty is not weakness
+- For controversial topics, present multiple perspectives fairly
+- Prefer primary sources over secondary sources over tertiary sources
+- When quoting, use exact text — do not paraphrase and present as a quote
+- If the user messages you mid-research, respond and then continue
+- Do not include sources you haven't actually read (no padding the bibliography)
+"""
+
+[dashboard]
+[[dashboard.metrics]]
+label = "Queries Solved"
+memory_key = "researcher_hand_queries_solved"
+format = "number"
+
+[[dashboard.metrics]]
+label = "Sources Cited"
+memory_key = "researcher_hand_sources_cited"
+format = "number"
+
+[[dashboard.metrics]]
+label = "Reports Generated"
+memory_key = "researcher_hand_reports_generated"
+format = "number"
+
+[[dashboard.metrics]]
+label = "Active Investigations"
+memory_key = "researcher_hand_active_investigations"
+format = "number"
--- a/crates/openfang-hands/bundled/researcher/SKILL.md
+++ b/crates/openfang-hands/bundled/researcher/SKILL.md
@@ -0,0 +1,327 @@
+---
+name: researcher-hand-skill
+version: "1.0.0"
+description: "Expert knowledge for AI deep research — methodology, source evaluation, search optimization, cross-referencing, synthesis, and citation formats"
+runtime: prompt_only
+---
+
+# Deep Research Expert Knowledge
+
+## Research Methodology
+
+### Research Process (5 phases)
+1. **Define**: Clarify the question, identify what's known vs unknown, set scope
+2. **Search**: Systematic multi-strategy search across diverse sources
+3. **Evaluate**: Assess source quality, extract relevant data, note limitations
+4. **Synthesize**: Combine findings into coherent answer, resolve contradictions
+5. **Verify**: Cross-check critical claims, identify remaining uncertainties
+
+### Question Types & Strategies
+| Question Type | Strategy | Example |
+|--------------|----------|---------|
+| Factual | Find authoritative primary source | "What is the population of Tokyo?" |
+| Comparative | Multi-source balanced analysis | "React vs Vue for large apps?" |
+| Causal | Evidence chain + counterfactuals | "Why did Theranos fail?" |
+| Predictive | Trend analysis + expert consensus | "Will quantum computing replace classical?" |
+| How-to | Step-by-step from practitioners | "How to set up a Kubernetes cluster?" |
+| Survey | Comprehensive landscape mapping | "What are the options for vector databases?" |
+| Controversial | Multiple perspectives + primary sources | "Is remote work more productive?" |
+
+### Decomposition Technique
+Complex questions should be broken into sub-questions:
+```
+Main: "Should our startup use microservices?"
+Sub-questions:
+  1. What are microservices? (definitional)
+  2. What are the benefits vs monolith? (comparative)
+  3. What team size/stage is appropriate? (contextual)
+  4. What are the operational costs? (factual)
+  5. What do similar startups use? (case studies)
+  6. What are the migration paths? (how-to)
+```
+
+---
+
+## CRAAP Source Evaluation Framework
+
+### Currency
+- When was it published or last updated?
+- Is the information still current for the topic?
+- Are the links functional?
+- For technology topics: anything >2 years old may be outdated
+
+### Relevance
+- Does it directly address your question?
+- Who is the intended audience?
+- Is the level of detail appropriate?
+- Would you cite this in your report?
+
+### Authority
+- Who is the author? What are their credentials?
+- What institution published this?
+- Is there contact information?
+- Does the URL domain indicate authority? (.gov, .edu, reputable org)
+
+### Accuracy
+- Is the information supported by evidence?
+- Has it been reviewed or refereed?
+- Can you verify the claims from other sources?
+- Are there factual errors, typos, or broken logic?
+
+### Purpose
+- Why does this information exist?
+- Is it informational, commercial, persuasive, or entertainment?
+- Is the bias clear or hidden?
+- Does the author/organization benefit from you believing this?
+
+### Scoring
+```
+A (Authoritative):  Passes all 5 CRAAP criteria
+B (Reliable):       Passes 4/5, minor concern on one
+C (Useful):         Passes 3/5, use with caveats
+D (Weak):           Passes 2/5 or fewer
+F (Unreliable):     Fails most criteria, do not cite
+```
+
+---
+
+## Search Query Optimization
+
+### Query Construction Techniques
+
+**Exact phrase**: `"specific phrase"` — use for names, quotes, error messages
+**Site-specific**: `site:domain.com query` — search within a specific site
+**Exclude**: `query -unwanted_term` — remove irrelevant results
+**File type**: `filetype:pdf query` — find specific document types
+**Recency**: `query after:2024-01-01` — recent results only
+**OR operator**: `query (option1 OR option2)` — broaden search
+**Wildcard**: `"how to * in python"` — fill-in-the-blank
+
+### Multi-Strategy Search Pattern
+For each research question, use at least 3 search strategies:
+1. **Direct**: The question as-is
+2. **Authoritative**: `site:gov OR site:edu OR site:org [topic]`
+3. **Academic**: `[topic] research paper [year]` or `site:arxiv.org [topic]`
+4. **Practical**: `[topic] guide` or `[topic] tutorial` or `[topic] how to`
+5. **Data**: `[topic] statistics` or `[topic] data [year]`
+6. **Contrarian**: `[topic] criticism` or `[topic] problems` or `[topic] myths`
+
+### Source Discovery by Domain
+| Domain | Best Sources | Search Pattern |
+|--------|-------------|---------------|
+| Technology | Official docs, GitHub, Stack Overflow, engineering blogs | `[tech] documentation`, `site:github.com [tech]` |
+| Science | PubMed, arXiv, Nature, Science | `site:arxiv.org [topic]`, `[topic] systematic review` |
+| Business | SEC filings, industry reports, HBR | `[company] 10-K`, `[industry] report [year]` |
+| Medicine | PubMed, WHO, CDC, Cochrane | `site:pubmed.ncbi.nlm.nih.gov [topic]` |
+| Legal | Court records, law reviews, statute databases | `[case] ruling`, `[law] analysis` |
+| Statistics | Census, BLS, World Bank, OECD | `site:data.worldbank.org [metric]` |
+| Current events | Reuters, AP, BBC, primary sources | `[event] statement`, `[event] official` |
+
+---
+
+## Cross-Referencing Techniques
+
+### Verification Levels
+```
+Level 1: Single source (unverified)
+  → Mark as "reported by [source]"
+
+Level 2: Two independent sources agree (corroborated)
+  → Mark as "confirmed by multiple sources"
+
+Level 3: Primary source + secondary confirmation (verified)
+  → Mark as "verified — primary source: [X]"
+
+Level 4: Expert consensus (well-established)
+  → Mark as "widely accepted" or "scientific consensus"
+```
+
+### Contradiction Resolution
+When sources disagree:
+1. Check which source is more authoritative (CRAAP scores)
+2. Check which is more recent (newer may have updated info)
+3. Check if they're measuring different things (apples vs oranges)
+4. Check for known biases or conflicts of interest
+5. Present both views with evidence for each
+6. State which view the evidence better supports (if clear)
+7. If genuinely uncertain, say so — don't force a conclusion
+
+---
+
+## Synthesis Patterns
+
+### Narrative Synthesis
+```
+The evidence suggests [main finding].
+
+[Source A] found that [finding 1], which is consistent with
+[Source B]'s observation that [finding 2]. However, [Source C]
+presents a contrasting view: [finding 3].
+
+The weight of evidence favors [conclusion] because [reasoning].
+A key limitation is [gap or uncertainty].
+```
+
+### Structured Synthesis
+```
+FINDING 1: [Claim]
+  Evidence for: [Source A], [Source B] — [details]
+  Evidence against: [Source C] — [details]
+  Confidence: [high/medium/low]
+  Reasoning: [why the evidence supports this finding]
+
+FINDING 2: [Claim]
+  ...
+```
+
+### Gap Analysis
+After synthesis, explicitly note:
+- What questions remain unanswered?
+- What data would strengthen the conclusions?
+- What are the limitations of the available sources?
+- What follow-up research would be valuable?
+
+---
+
+## Citation Formats
+
+### Inline URL
+```
+According to a 2024 study (https://example.com/study), the effect was significant.
+```
+
+### Footnotes
+```
+According to a 2024 study[1], the effect was significant.
+
+---
+[1] https://example.com/study — "Title of Study" by Author, Published Date
+```
+
+### Academic (APA)
+```
+In-text: (Smith, 2024)
+Reference: Smith, J. (2024). Title of the article. *Journal Name*, 42(3), 123-145. https://doi.org/10.xxxx
+```
+
+For web sources (APA):
+```
+Author, A. A. (Year, Month Day). Title of page. Site Name. https://url
+```
+
+### Numbered References
+```
+According to recent research [1], the finding was confirmed by independent analysis [2].
+
+## References
+1. Author (Year). Title. URL
+2. Author (Year). Title. URL
+```
+
+---
+
+## Output Templates
+
+### Brief Report
+```markdown
+# [Question]
+**Date**: YYYY-MM-DD | **Sources**: N | **Confidence**: high/medium/low
+
+## Answer
+[2-3 paragraph direct answer]
+
+## Key Evidence
+- [Finding 1] — [source]
+- [Finding 2] — [source]
+- [Finding 3] — [source]
+
+## Caveats
+- [Limitation or uncertainty]
+
+## Sources
+1. [Source](url)
+2. [Source](url)
+```
+
+### Detailed Report
+```markdown
+# Research Report: [Question]
+**Date**: YYYY-MM-DD | **Depth**: thorough | **Sources Consulted**: N
+
+## Executive Summary
+[1 paragraph synthesis]
+
+## Background
+[Context needed to understand the findings]
+
+## Methodology
+[How the research was conducted, what was searched, how sources were evaluated]
+
+## Findings
+
+### [Sub-question 1]
+[Detailed findings with inline citations]
+
+### [Sub-question 2]
+[Detailed findings with inline citations]
+
+## Analysis
+[Synthesis across findings, patterns identified, implications]
+
+## Contradictions & Open Questions
+[Areas of disagreement, gaps in knowledge]
+
+## Confidence Assessment
+[Overall confidence level with reasoning]
+
+## Sources
+[Full bibliography in chosen citation format]
+```
+
+---
+
+## Cognitive Bias in Research
+
+Be aware of these biases during research:
+
+1. **Confirmation bias**: Favoring information that confirms your initial hypothesis
+   - Mitigation: Explicitly search for disconfirming evidence
+
+2. **Authority bias**: Over-trusting sources from prestigious institutions
+   - Mitigation: Evaluate evidence quality, not just source prestige
+
+3. **Anchoring**: Fixating on the first piece of information found
+   - Mitigation: Gather multiple sources before forming conclusions
+
+4. **Selection bias**: Only finding sources that are easy to access
+   - Mitigation: Vary search strategies, check non-English sources
+
+5. **Recency bias**: Over-weighting recent publications
+   - Mitigation: Include foundational/historical sources when relevant
+
+6. **Framing effect**: Being influenced by how information is presented
+   - Mitigation: Look at raw data, not just interpretations
+
+---
+
+## Domain-Specific Research Tips
+
+### Technology Research
+- Always check the official documentation first
+- Compare documentation version with the latest release
+- Stack Overflow answers may be outdated — check the date
+- GitHub issues/discussions often have the most current information
+- Benchmarks without methodology descriptions are unreliable
+
+### Business Research
+- SEC filings (10-K, 10-Q) are the most reliable public company data
+- Press releases are marketing — verify claims independently
+- Analyst reports may have conflicts of interest — check disclaimers
+- Employee reviews (Glassdoor) provide internal perspective but are biased
+
+### Scientific Research
+- Systematic reviews and meta-analyses are strongest evidence
+- Single studies should not be treated as definitive
+- Check if findings have been replicated
+- Preprints have not been peer-reviewed — note this caveat
+- p-values and effect sizes both matter — not just "statistically significant"