Data & analysis

Chatmosp Literature Search

Try it

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.

What it does

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.

The skill document

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.md in this same directory and respond in Chinese. Always match the response language to the user's input language.

1. Core Responsibilities

  1. Journal search: access 3 priority layers
  2. Article retrieval: 3 stages (abstract → main text → SI)
  3. Parameter extraction: extract E_ads, w, gamma from tables
  4. Parameter validation: completeness + reasonability + consistency
  5. Return scored result to parameter-builder

2. Input Contract

FieldRequiredExample
Metal elementYesPd, Pt, Au, Cu, Ni
Gas systemYesCO+O₂, H₂+CO₂, CO+H₂
Temperature rangeNo300-500K
Pressure rangeNo1-10 atm

3. Output Contract

Return parameter table (JSON), containing:

FieldDescription
Surface energyγ (eV/Ų) for each facet (100, 110, 111)
Adsorption energyE_ads (eV) for each facet + each gas
Interaction matrixw (eV) for CO-CO, CO-O, O-O
Parameter sourceDOI, article title, authors
Completeness score1-10

⚠️ Literature search does NOT return gas entropy. parameter-builder MUST recalculate it using the §3.1 formula after receiving the result.

4. Workflow

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)

5. Journal Search Priority

Layer 1: Top journals (priority, but may be inaccessible)

  • Science, Nature, JACS, Angewandte Chemie, PRL, JCP
  • Strategy: DOI resolution → abstract → main text → SI
  • API limit or CAPTCHA → immediately skip to Layer 2
  • Nature Communications
  • Science Advances
  • PNAS
  • ACS Central Science
  • Chemical Science

Advantages: completely free, high quality, complete SI

Layer 3: Preprint platforms (last resort)

  • arXiv, ChemRxiv, bioRxiv
  • Note: not peer-reviewed, parameters may be inaccurate, requires user confirmation

6. Tool Selection

6.1 Prefer the currently available interactive browser/web tool

  • ✅ Supports JS rendering
  • ✅ Supports login state
  • ✅ Supports interactive operations
  • ✅ Better suited for journal sites

6.2 Do NOT use web_fetch

  • ❌ Journal sites often have restrictions (403, CAPTCHA)
  • ❌ Cannot execute JS
  • ❌ Cannot get dynamically loaded content

6.3 Fallback

If 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.

7. Detailed Literature Retrieval Process

Step 1: Build search keywords

✅ Correct format (MUST use):

"keyword1" AND "keyword2" AND "keyword3"

Examples:

  • Cu cluster in CO oxidation → "Cu" AND "CO oxidation"
  • Pd cluster in CO oxidation → "Pd" AND "CO oxidation"
  • Pt cluster in H₂ → "Pt" AND "H2"

❌ Incorrect format (DO NOT use):

Pd CO oxidation

Space-separated format is treated as one keyword, returning inaccurate results.

Step 2: Access journal site

  • Use the currently available browser/web tool to enter journal search page
  • Input keywords
  • Default sort: relevance

Step 3: Get top 10 results

  • Only check top 10 (highest relevance)
  • Output title, DOI, abstract (optional)

Step 4: Initial screening (by title)

Screening criteria:

  1. Reaction atmosphere match (must be target reaction: CO oxidation, WGSR, etc.)
  2. Metal system match (must be target metal: Pd, Pt, Cu, etc.; exclude alloys)

Output filtered list + screening reason for each

Step 5: User interaction

Choose next step:

  1. Carefully check all filtered articles
  2. Carefully check one specific article (user-specified)
  3. Switch to another journal

Step 6: Detailed article check

6.1: Check main text

  • Confirm metal match
  • Confirm reaction atmosphere match
  • Look for parameter tables

6.2: No parameters → download SI

  • Parameters only exist in text and tables, NOT in images
  • Download SI PDF: si_{first_author}_{year}.pdf
  • Same author, same year, multiple: si_smith_2023a.pdf / b / c

6.3: Check SI

  • Search keywords: Table, Supplementary Table, S1, S2
  • Extract parameter tables

Step 7: Parameter extraction

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

  • Table S, Supplementary Table
  • adsorption energy, E_ads, binding energy
  • surface energy, γ, surface tension
  • interaction parameter, interaction matrix

7.4: Use read tool to read text, locate parameter tables

7.5: Extract values from tables, note unit conversions

7.6: Validate parameters

  • Within reasonable range
  • Follow physical laws
  • Consistent

Step 8: Show parameters to user

📊 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

8. Article Retrieval 3 Stages

StagePurposeSearch content
1. AbstractConfirm relevanceMetal, reaction, research type
2. Main textFind parameter tablesMethods, Results, Tables, Figures
3. SIComplete parametersSupplementary 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.

9. MSR Required Data Checklist

Basic parameters

  • ✅ Metal element, temperature, pressure
  • ✅ Cluster radius, gas species, gas partial pressures

Surface parameters (per facet)

  • ✅ Surface energy γ (eV/Ų): (100), (110), (111), (211), (311), etc.

Adsorption parameters (per facet + per gas)

  • ✅ Adsorption energy E_ads (eV): e.g., CO on Pd(111), O₂ on Pd(100)

Interaction parameters

  • ✅ w matrix (eV): CO-CO, CO-O, O-O

Gas parameters

  • ✅ Gas entropy (eV/K): can be calculated via formula (T-dependent) or extracted

⚠️ Literature search does NOT return gas entropy; parameter-builder MUST auto-calculate

10. Completeness Scoring

Per-parameter score

Parameter typeScoreDescription
Surface energy2 pointsAt least (100), (110), (111) three facets
Adsorption energy3 pointsEach gas on each facet
Interaction matrix3 pointsCO-CO, CO-O, O-O
Parameter source1 pointDOI, title, author
Parameter reasonability1 pointWithin reasonable range, physically plausible

Score level action

ScoreLevelAction
9-10CompleteUse directly
7-8Mostly completeUser confirms missing, then use
5-6PartialSupplement missing
3-4IncompleteUse similar metal as reference
1-2Very incompleteNot recommended

⚠️ Gas entropy is NOT in this scoring. Regardless of score, parameter-builder MUST recalculate gas entropy per formula.

11. Timeout & Fallback

11.1 Timeout

  • Total time limit: 5 minutes
  • Per-article limit: 2 minutes
  • On timeout: stop immediately, return found parameters (even if incomplete)

11.2 Fallback

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

11.3 One journal failure prompt

⚠️ 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)

12. Lessons Learned

  • Lesson 1: Keyword selection
    • Start with broad keywords ("Pd" AND "CO oxidation")
    • Then filter with specific ("adsorption energy")
    • Search topics, not just parameter names
  • Lesson 2: DOI priority
    • Prefer known DOI or title
    • Direct DOI access
    • Don't rely solely on keyword search
  • Lesson 3: Multi-platform search
    • Don't limit to single journal
    • Prefer open access
  • Lesson 4: Don't waste time on inaccessible resources
    • API limit or CAPTCHA → immediately skip to next layer
    • Prioritize open access
  • Lesson 5: Main text not mentioning parameters ≠ SI lacks them
    • If article is highly relevant, still download SI
    • SI typically has complete parameter tables

13. Cross-Skill Handoff

  • Called by: parameter-builder when key parameters missing + user chooses lit search
  • Calls tools: available browser/web/PDF/file-reading tools in the current environment
  • Returns to parameter-builder: parameter table with score

14. Dependencies

  • chatmosp-parameter-builder — caller
  • Browser/web tool — journal site access, SI download
  • pdftotext — PDF to text
  • read — read text files

15. File Structure

chatmosp-literature-search/
├── SKILL.md       # Chinese reference
└── SKILL_cn.md    # This file (English)

16. Practical Example

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)

ChatMOSP Companion Skills

This skill is part of the ChatMOSP skill set. For complete functionality, install the companion ChatMOSP skills from ClawHub when they are missing:

  • chatmosp-input-coordinator
  • chatmosp-parameter-builder
  • chatmosp-msr-generator
  • chatmosp-kmc-simulator
  • chatmosp-literature-search
  • chatmosp-file-organizer

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.

Related skills

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

1 installs

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

1 installs

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.

1 installs

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

2 installs

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

2 installs

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

1 stars