UNPKG

aiwg

Version:

Deployment tool and support utility for AI context. Copies agents, skills, commands, rules, and behaviors into the paths each AI platform reads (Claude Code, Codex, Copilot, Cursor, Warp, OpenClaw, and 6 more) so one source of truth works across 10 platfo

466 lines (333 loc) 14.2 kB
# Quality Assessment Report: REF-XXX ## Metadata ```yaml --- assessment_type: source-quality ref_id: REF-XXX source_title: "[Paper Title]" assessment_date: YYYY-MM-DD assessor: quality-agent # or manual-expert assessor_version: 1.0.0 overall_score: 0 # 0-100 overall_grade: null # High Quality (70-100), Moderate (50-69), Low (<50) grade_rating: null # GRADE: High, Moderate, Low, Very Low fair_compliant: false # true/false fair_score: 0 # 0-4 (number of FAIR principles met) assessment_status: draft # draft | reviewed | final --- ``` ## Executive Summary **Overall Quality Score:** X / 100 **Quality Grade:** [High Quality | Moderate Quality | Low Quality] **GRADE Rating:** [High | Moderate | Low | Very Low] **FAIR Compliance:** [X/4 principles met] **Recommendation:** [Approved for integration | Conditional approval | Seek alternative source] **Summary:** [2-3 sentence assessment: Is this source suitable for AIWG research? What are the key strengths/limitations?] --- ## Multi-Dimensional Quality Assessment ### 1. Authority Evaluation (Weight: 30%) **Score:** X / 100 #### Author Credentials | Criterion | Assessment | Evidence | Points | |-----------|------------|----------|--------| | Academic Affiliation | [Yes/No/Unknown] | [Institution name or "None"] | X/25 | | Subject Expertise | [Expert/Intermediate/Novice] | [Publication history, h-index] | X/25 | | Publication History | [Extensive/>10 papers/Few/<5/None] | [Google Scholar profile] | X/25 | | Industry Recognition | [Award/Conference speaker/Standard/None] | [Specific awards or roles] | X/25 | **Author Score:** X / 100 #### Publisher Reputation | Criterion | Assessment | Evidence | Points | |-----------|------------|----------|--------| | Venue Tier | [A*/A/B/C/Blog/Unknown] | [CORE ranking or equivalent] | X/30 | | Peer Review Status | [Peer-reviewed/Preprint/Blog/None] | [Publication type] | X/30 | | Publisher Prestige | [Top-tier/Reputable/Standard/Unknown] | [Journal impact factor, conference acceptance rate] | X/25 | | Citation Count | [>1000/100-1000/10-100/<10] | [Semantic Scholar, Google Scholar] | X/15 | **Publisher Score:** X / 100 **Authority Overall:** (Author Score × 0.5) + (Publisher Score × 0.5) = **X / 100** **Rationale:** [Why this authority score? What are the key factors?] --- ### 2. Currency Evaluation (Weight: 20%) **Score:** X / 100 #### Timeliness | Criterion | Assessment | Evidence | Points | |-----------|------------|----------|--------| | Publication Date | [<1 year/1-2 years/3-5 years/>5 years] | YYYY | X/40 | | Topic Currency | [Active research/Established/Declining] | [Publication trend analysis] | X/30 | | Access Date | [Recent/<6 months/6-12 months/>1 year] | YYYY-MM-DD | X/15 | | Updates Available | [Yes/No/N/A] | [Revised version, errata] | X/15 | **Currency Overall:** X / 100 **Rationale:** [Is this source still current for the topic? Has research moved on?] **Field Context:** - **Fast-Moving Field** (LLMs, AI): <2 years = current, >3 years = dated - **Established Field** (SDLC): <5 years = current, >10 years = dated --- ### 3. Accuracy Evaluation (Weight: 25%) **Score:** X / 100 #### Methodological Rigor | Criterion | Assessment | Evidence | Points | |-----------|------------|----------|--------| | Peer Review | [Yes/Preprint/None] | [Journal/conference peer review] | X/30 | | Methodology Documented | [Comprehensive/Adequate/Minimal/None] | [Methods section quality] | X/25 | | Data Sources Cited | [All cited/Most cited/Some/None] | [Reference quality] | X/20 | | Reproducibility | [Fully reproducible/Partially/Not reproducible] | [Code/data availability] | X/15 | | Conflicts of Interest | [Disclosed/None/Undisclosed] | [COI statement] | X/10 | **Accuracy Overall:** X / 100 **Rationale:** [How rigorous is the research? Can results be trusted?] **Red Flags (if any):** - [ ] No methodology described - [ ] Data sources unclear - [ ] Conflicts of interest not disclosed - [ ] Extraordinary claims without extraordinary evidence --- ### 4. Coverage Evaluation (Weight: 15%) **Score:** X / 100 #### Comprehensiveness | Criterion | Assessment | Evidence | Points | |-----------|------------|----------|--------| | Topic Depth | [Comprehensive/Adequate/Superficial] | [Analysis detail level] | X/35 | | Breadth of Analysis | [Multi-faceted/Focused/Narrow] | [Perspectives covered] | X/30 | | Completeness | [Thorough/Adequate/Incomplete] | [Key aspects addressed] | X/25 | | Related Work Coverage | [Extensive/Adequate/Minimal/None] | [Literature review quality] | X/10 | **Coverage Overall:** X / 100 **Rationale:** [Does this source provide sufficient depth and breadth?] --- ### 5. Objectivity Evaluation (Weight: 10%) **Score:** X / 100 #### Bias Assessment | Criterion | Assessment | Evidence | Points | |-----------|------------|----------|--------| | Bias Detection | [Minimal/Moderate/Significant] | [Subjective language, cherry-picking] | X/40 | | Balanced Perspective | [Yes/Mostly/No] | [Multiple viewpoints presented] | X/30 | | Conflicts of Interest | [None/Disclosed/Undisclosed] | [Financial, affiliation] | X/20 | | Objectivity Language | [Neutral/Mostly neutral/Subjective] | [Tone analysis] | X/10 | **Objectivity Overall:** X / 100 **Rationale:** [Are there bias concerns? Is the perspective balanced?] **Potential Biases:** - [ ] Corporate affiliation (vendor bias) - [ ] Financial conflict of interest - [ ] Confirmation bias (cherry-picked evidence) - [ ] Publication bias (only positive results) --- ## Overall Quality Score Calculation | Dimension | Weight | Score | Weighted Score | |-----------|--------|-------|----------------| | **Authority** | 30% | X/100 | X.X | | **Currency** | 20% | X/100 | X.X | | **Accuracy** | 25% | X/100 | X.X | | **Coverage** | 15% | X/100 | X.X | | **Objectivity** | 10% | X/100 | X.X | **Overall Quality Score:** X / 100 **Quality Grade:** - **High Quality (70-100):** Approved for integration, suitable for critical decisions - **Moderate Quality (50-69):** Conditional approval, verify with additional sources - **Low Quality (<50):** Not recommended, seek higher-quality alternatives --- ## GRADE Assessment (Evidence Quality) ### GRADE Dimensions #### 1. Risk of Bias (Study Design Quality) **Score:** X / 25 points | Study Design | Starting GRADE | Score | |--------------|----------------|-------| | Randomized Controlled Trial (RCT) | High | 25 | | Well-designed observational study | Moderate | 20 | | Observational study | Low | 15 | | Case study / Expert opinion | Very Low | 5 | **This Source:** [Study design type] **X points** **Bias Assessment:** - Selection bias: [Low/Moderate/High] - Performance bias: [Low/Moderate/High] - Detection bias: [Low/Moderate/High] **Adjustment:** [No downgrade | Downgrade by X points due to Y bias] --- #### 2. Consistency (Agreement with Other Studies) **Score:** X / 20 points | Consistency Level | Assessment | Score | |-------------------|------------|-------| | Aligns with other studies | No issues | 20 | | Mostly consistent | Minor inconsistencies | 15 | | Contradicts some studies | Conflicting evidence | 10 | | Unknown (first study) | Cannot assess | 10 | **This Source:** [Assessment] **X points** **Evidence:** [Compare findings to REF-YYY, REF-ZZZ - do they agree or contradict?] --- #### 3. Directness (Relevance to Question) **Score:** X / 20 points | Directness Level | Assessment | Score | |------------------|------------|-------| | Directly addresses question | Direct | 20 | | Addresses similar question | Mostly direct | 15 | | Indirectly relevant | Indirect | 10 | | Tangential relevance | Very indirect | 5 | **This Source:** [Assessment] **X points** **Rationale:** [How directly does this address AIWG research question?] --- #### 4. Precision (Statistical Power) **Score:** X / 20 points | Precision Level | Assessment | Score | |-----------------|------------|-------| | Large sample, narrow CI | High precision | 20 | | Adequate sample | Moderate precision | 15 | | Small sample, wide CI | Low precision | 10 | | Single case study | Very low precision | 5 | **This Source:** - **Sample Size:** n=X - **Confidence Interval:** [95% CI: X-Y] or [Not reported] - **Statistical Power:** [Adequate/Inadequate] **Score:** X points --- #### 5. Publication Bias (Selective Reporting) **Score:** X / 15 points | Venue Tier | Assessment | Score | |------------|------------|-------| | A* venue (top-tier) | Minimal bias | 15 | | A venue (reputable) | Low bias | 12 | | B venue (standard) | Moderate bias | 8 | | C venue or preprint | Higher bias | 5 | | Blog/unreviewed | Significant bias | 3 | **This Source:** [Venue tier] **X points** **Publication Bias Assessment:** - [ ] Negative results reported (reduces bias) - [ ] Industry-funded (may increase bias) - [ ] Preprint (not peer-reviewed) --- ### GRADE Overall Score **Total GRADE Score:** X / 100 points **GRADE Rating:** - **High (80-100):** Strong confidence in evidence, suitable for critical decisions - **Moderate (60-79):** Moderate confidence, likely suitable but verify - **Low (40-59):** Limited confidence, use with caution - **Very Low (<40):** Very limited confidence, seek better sources **Confidence Statement:** [We have {High/Moderate/Low/Very Low} confidence that this evidence is reliable for informing AIWG development decisions.] --- ## FAIR Principles Validation ### 1. Findable **Status:** [ Met | Not Met] **Checklist:** - [ ] **Persistent Identifier:** DOI, ArXiv ID, ISBN, or Handle - **Evidence:** [DOI or identifier] - [ ] **Metadata Complete:** Title, authors, year, abstract documented - **Evidence:** [Metadata file exists] - [ ] **Registered in Index:** Searchable in research corpus - **Evidence:** [REF-XXX indexed] **Score:** X / 3 criteria met **Rationale:** [Why met or not met?] --- ### 2. Accessible **Status:** [ Met | Not Met] **Checklist:** - [ ] **Retrieval Protocol:** URL or API documented - **Evidence:** [URL or access method] - [ ] **Access Conditions Clear:** Open access, subscription, or manual upload - **Evidence:** [License type] - [ ] **Long-Term Storage:** Archival plan or stable repository - **Evidence:** [Publisher commitment, local backup] **Score:** X / 3 criteria met **Rationale:** [Why met or not met?] --- ### 3. Interoperable **Status:** [ Met | Not Met] **Checklist:** - [ ] **Standard Format:** PDF/A, BibTeX, or machine-readable format - **Evidence:** [File format] - [ ] **Controlled Vocabulary:** Tags use standard ontology - **Evidence:** [Tag schema] - [ ] **Linked Data:** Cross-references to related sources - **Evidence:** [Links to REF-YYY, REF-ZZZ] **Score:** X / 3 criteria met **Rationale:** [Why met or not met?] --- ### 4. Reusable **Status:** [ Met | Not Met] **Checklist:** - [ ] **Clear License:** CC-BY, MIT, proprietary explicitly stated - **Evidence:** [License info] - [ ] **Provenance Documented:** Acquisition and documentation history - **Evidence:** [Provenance log exists] - [ ] **Usage Context:** How to use this source documented - **Evidence:** [README or usage notes] **Score:** X / 3 criteria met **Rationale:** [Why met or not met?] --- ### FAIR Overall Compliance **FAIR Score:** X / 4 principles met **Compliance Level:** - **Fully Compliant (4/4):** Archival-ready, suitable for long-term preservation - **Partially Compliant (2-3/4):** Acceptable with documentation of limitations - **Non-Compliant (0-1/4):** Remediate or exclude from archival packages **Remediation Recommendations (if <4/4):** 1. [Action to improve Findable] 2. [Action to improve Accessible] 3. [Action to improve Interoperable] 4. [Action to improve Reusable] --- ## Strengths and Limitations ### Strengths 1. **[Strength 1]:** [Explanation] 2. **[Strength 2]:** [Explanation] 3. **[Strength 3]:** [Explanation] ### Limitations 1. **[Limitation 1]:** [Explanation and impact] 2. **[Limitation 2]:** [Explanation and impact] 3. **[Limitation 3]:** [Explanation and impact] ### Critical Issues (if any) - [ ] **Issue 1:** [Description and why it's critical] - [ ] **Issue 2:** [Description and why it's critical] --- ## Recommendations ### Usage Recommendation **Approved for Integration:** [Yes | Conditional | No] **Conditions (if conditional):** 1. [Condition 1 that must be met] 2. [Condition 2 that must be met] **Rationale:** [Why is this source suitable or unsuitable for AIWG research corpus?] ### Citation Confidence **Confidence Level:** [High | Moderate | Low] **Guidance:** - **High Confidence:** Use for critical architectural decisions, cite prominently - **Moderate Confidence:** Use for supporting evidence, cite with context - **Low Confidence:** Reference for awareness only, do not rely on for decisions ### Alternative Sources (if Low Quality) [If quality score <50, recommend better alternatives] 1. **Alternative 1:** [REF-YYY or search query] 2. **Alternative 2:** [REF-ZZZ or search query] --- ## Validation Rules ### Required Fields - `ref_id`: Valid REF-XXX identifier - `assessment_date`: ISO 8601 date - `overall_score`: Integer 0-100 - `grade_rating`: Valid GRADE rating - `fair_score`: Integer 0-4 ### Score Consistency - Overall score must equal weighted average of dimension scores - GRADE score must follow dimension scoring rules - FAIR score must match number of principles met --- ## Agent Responsibilities **Produced by:** Quality Agent (UC-RF-006) **Updated by:** Quality Agent (periodic reassessment), Manual Expert (validation) **Used by:** Citation Agent (confidence levels), Gap Analysis Agent (corpus quality), User (source selection) --- ## References - @.aiwg/research/sources/metadata/REF-XXX-metadata.json - Source metadata - @$AIWG_ROOT/agentic/code/frameworks/research-complete/elaboration/use-cases/UC-RF-006-assess-source-quality.md - Quality assessment use case - [GRADE Framework](https://www.gradeworkinggroup.org/) - Evidence grading standard - [FAIR Principles](https://www.go-fair.org/fair-principles/) - Data quality standard --- **Template Version:** 1.0.0 **Last Updated:** 2026-01-25 **Owner:** Quality Agent