数据分析
Academic literature searcher of the chatMOSP system. When MOSP_database lacks matching parameters, searches open-access journals (Nature Communications, Science Advances, PNAS, etc.) and other academic resources to extract MSR/KMC parameters.
Academic literature searcher of the chatMOSP system. When MOSP_database lacks matching parameters, searches open-access journals (Nature Communications, Science Advances, PNAS, etc.) and other academic resources to extract MSR/KMC parameters (surface energy, adsorption energy, interaction matrix, etc.). Triggers: parameter-builder detects missing key parameters and the user chooses literature search for completion.
Language routing / 语言路由 Detect the user's language from the latest message.
- If the user writes in English, continue with this English file and respond in English.
- If the user writes in Chinese, read
SKILL.mdin this same directory and respond in Chinese. Always match the response language to the user's input language.
| Field | Required | Example |
|---|---|---|
| Metal element | Yes | Pd, Pt, Au, Cu, Ni |
| Gas system | Yes | CO+O₂, H₂+CO₂, CO+H₂ |
| Temperature range | No | 300-500K |
| Pressure range | No | 1-10 atm |
Return parameter table (JSON), containing:
| Field | Description |
|---|---|
| Surface energy | γ (eV/Ų) for each facet (100, 110, 111) |
| Adsorption energy | E_ads (eV) for each facet + each gas |
| Interaction matrix | w (eV) for CO-CO, CO-O, O-O |
| Parameter source | DOI, article title, authors |
| Completeness score | 1-10 |
⚠️ Literature search does NOT return gas entropy. parameter-builder MUST recalculate it using the §3.1 formula after receiving the result.
Input (metal + gas)
→ Journal search (3 priority layers)
→ Article retrieval (3 stages: abstract → main text → SI)
→ Parameter extraction (pdftotext + keyword search)
→ Parameter validation (completeness, reasonability, consistency)
→ Output (parameter table with score)
Advantages: completely free, high quality, complete SI
web_fetchIf the current environment has no browser, web-search, PDF, or file-reading tool available, do not pretend that literature retrieval was completed. Ask the user to provide the paper, DOI, SI file, or parameter values directly.
✅ Correct format (MUST use):
"keyword1" AND "keyword2" AND "keyword3"
Examples:
"Cu" AND "CO oxidation""Pd" AND "CO oxidation""Pt" AND "H2"❌ Incorrect format (DO NOT use):
Pd CO oxidation
Space-separated format is treated as one keyword, returning inaccurate results.
Screening criteria:
Output filtered list + screening reason for each
Choose next step:
6.1: Check main text
6.2: No parameters → download SI
si_{first_author}_{year}.pdfsi_smith_2023a.pdf / b / c6.3: Check SI
7.1: Download SI via the currently available browser/web tool
7.2: Convert PDF to text
pdftotext si_{author}_{year}.pdf si_{author}_{year}.txt
7.3: Search parameter table keywords
7.4: Use read tool to read text, locate parameter tables
7.5: Extract values from tables, note unit conversions
7.6: Validate parameters
📊 Parameters extracted from literature:
【Article Information】
- Title: {title}
- DOI: {doi}
- Journal: {journal}
【Parameter Table】
- Surface energy: {values}
- Adsorption energy: {values}
- Interaction matrix: {values}
【Completeness Score】
- Score: {score}/10
- Description: {description}
Please select:
1. ✅ Use these parameters → pass to parameter-builder
2. ❌ Reject, continue searching
3. ❌ Cancel task
| Stage | Purpose | Search content |
|---|---|---|
| 1. Abstract | Confirm relevance | Metal, reaction, research type |
| 2. Main text | Find parameter tables | Methods, Results, Tables, Figures |
| 3. SI | Complete parameters | Supplementary Tables, Methods, Data |
⚠️ IMPORTANT: Main text not mentioning parameters does NOT mean SI lacks them. If article relevance is high, still download SI to check.
⚠️ Literature search does NOT return gas entropy; parameter-builder MUST auto-calculate
| Parameter type | Score | Description |
|---|---|---|
| Surface energy | 2 points | At least (100), (110), (111) three facets |
| Adsorption energy | 3 points | Each gas on each facet |
| Interaction matrix | 3 points | CO-CO, CO-O, O-O |
| Parameter source | 1 point | DOI, title, author |
| Parameter reasonability | 1 point | Within reasonable range, physically plausible |
| Score | Level | Action |
|---|---|---|
| 9-10 | Complete | Use directly |
| 7-8 | Mostly complete | User confirms missing, then use |
| 5-6 | Partial | Supplement missing |
| 3-4 | Incomplete | Use similar metal as reference |
| 1-2 | Very incomplete | Not recommended |
⚠️ Gas entropy is NOT in this scoring. Regardless of score, parameter-builder MUST recalculate gas entropy per formula.
After 5 articles without complete parameters:
⚠️ I have searched 5 articles but could not find complete parameters.
- Surface energy: ✅ found
- Adsorption energy: ❌ not found
- Interaction matrix: ❌ not found
Suggestions:
1. Use similar metal as reference
2. Provide known parameter source (DOI or title)
3. Use default/test parameters
⚠️ No parameters found in top 10 articles of {Journal}
Please select:
1. Switch journal
2. Cancel task (if all journals exhausted, suggest changing system or user-providing params)
"Pd" AND "CO oxidation")"adsorption energy")chatmosp-literature-search/
├── SKILL.md # Chinese reference
└── SKILL_cn.md # This file (English)
User: show me the Pd cluster under atmosphere of CO and O2
1. Extract need: metal=Pd, gases=[CO, O₂]
2. Check MOSP_database: no match
3. parameter-builder prompts with 4 options → user chooses literature search (open access)
4. Launch this skill:
- Platform: Nature Communications
- Keywords: `"Pd" AND "CO oxidation" AND "adsorption energy"`
- Tool: currently available browser/web tool
5. Find relevant article, confirm DOI
6. Download SI → save to literature/
7. pdftotext → search "Supplementary Table" → extract parameters
8. Validate: completeness score
9. Display result + annotate missing
10. User confirms → pass to parameter-builder to assemble input.json
Key points:
- Always start with open-access journals
- Literature search does NOT return gas entropy; parameter-builder MUST recalculate
- Interaction parameters may be in KMC format; need MSR/KMC format conversion (see parameter-builder §9)
This skill is part of the ChatMOSP skill set. For complete functionality, install the companion ChatMOSP skills from ClawHub when they are missing:
If a required companion skill is unavailable, ask the user to install the missing ChatMOSP skill from ClawHub before continuing. Use the GitHub repository only as a fallback.
Parameter building and management center of the chatMOSP system. Handles MSR / KMC parameter queries, intelligent completion, gas entropy calculation, MSR↔KMC interaction parameter conversion, and missing-parameter handling. Triggers: after input-coordinator identifies a task but before the calculat
Entry skill of the chatMOSP system. Parses Chinese/English natural language input, recognizes three task types (MSR / KMC / parameter query), extracts metal, temperature, gas, partial pressure, steps, and size parameters, then dispatches parameter-builder, file-organizer, msr-generator, and kmc-simu
KMC (Kinetic Monte Carlo) simulation engine of the chatMOSP system. Invokes kmc_standalone.py via Wine to run the Windows main.exe engine, executes catalyst surface reaction kinetic simulations, and produces TOF / coverage results.
MSR (Multiscale Structure Reconstruction) calculation engine of the chatMOSP system. Invokes mosp-for-chatMOSP/utils/msr.py to generate metal cluster structures (Wulff construction), producing ini.xyz and {task_name}_cluster.xyz, then automatically generates PNG structure images and GIF rotation ani
File system manager of the chatMOSP system. Handles MSR / KMC task directory creation, intelligent naming (in `metal_gas-pp_T_K_P_Pa_R_size` format), and secure file operations (path-traversal protection, whitelist restriction). Triggers: after parameter-builder confirms parameters and before calcul
Catalysis literature search skill. Given a user-provided catalyst topic, reaction type, or material system, retrieve and organize relevant catalysis research papers from the open web (ScienceDirect, arXiv, OpenAlex, Google Scholar, etc.) and output structured literature-review information (literatur