[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M73J212GPVR6D33NEX2B4CQS":3,"skill-same:01M73J212GPVR6D33NEX2B4CQS":48},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M73J212GPVR6D33NEX2B4CQS","pytdc-therapeutics-data-commons",null,"Therapeutics Data Commons (TDC) AI-ready drug discovery datasets. Curated ADME, toxicity, DTI, DDI with scaffold\u002Fcold splits, standardized metrics, molecular oracles, and ADMET benchmarks for therapeutic ML and property prediction. For chemical database queries use chembl-database-bioactivity; for featurization use molfeat.","bg-szy\u002FTOP-SKILLS","bg-szy","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F106576190?v=4","01M73A4E8KPV67SFXRY8XZGNBX","ShuZhiyong","01M73A4E96MQ294MHM74TZBVQD",6,"MIT","https:\u002F\u002Fgithub.com\u002Fbg-szy\u002FTOP-SKILLS\u002Fblob\u002Fmaster\u002Fskills\u002Fawesome-skills\u002Fpytdc-therapeutics-data-commons\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: pytdc-therapeutics-data-commons\ndescription: \"Therapeutics Data Commons (TDC) AI-ready drug discovery datasets. Curated ADME, toxicity, DTI, DDI with scaffold\u002Fcold splits, standardized metrics, molecular oracles, and ADMET benchmarks for therapeutic ML and property prediction. For chemical database queries use chembl-database-bioactivity; for featurization use molfeat.\"\nlicense: MIT\n---\n\n# PyTDC (Therapeutics Data Commons)\n\n## Overview\n\nPyTDC is an open-science platform providing AI-ready datasets and benchmarks for drug discovery. It organizes therapeutics data into three categories: single-instance prediction (molecular\u002Fprotein properties), multi-instance prediction (drug-target interactions), and generation (molecule design, retrosynthesis). All datasets come with standardized splits, evaluation metrics, and molecular oracles.\n\n## When to Use\n\n- Loading curated ADME, toxicity, or bioactivity datasets for ML model training\n- Benchmarking drug discovery models with standardized 5-seed evaluation protocols\n- Predicting drug-target or drug-drug interactions with proper cold-split evaluation\n- Generating novel molecules and scoring them with molecular oracles (QED, SA, DRD2, GSK3B)\n- Accessing scaffold-based or temporal train\u002Ftest splits for pharmaceutical ML\n- Converting molecular representations (SMILES to PyG graphs, ECFP fingerprints, SELFIES)\n- For chemical database queries (compound search, bioactivity), use `chembl-database-bioactivity` instead\n- For molecular featurization beyond format conversion, use `molfeat` instead\n\n## Prerequisites\n\n```bash\nuv pip install PyTDC\n# Core deps: numpy, pandas, scikit-learn, tqdm, fuzzywuzzy\n# Optional: rdkit (scaffold splits), torch-geometric (PyG conversion)\n```\n\n**API Note**: TDC downloads datasets on first access (~10-500 MB per dataset). Specify `path='data\u002F'` to control download location. No API key required.\n\n## Quick Start\n\n```python\nfrom tdc.single_pred import ADME\nfrom tdc import Evaluator\n\n# Load dataset with scaffold split\ndata = ADME(name='Caco2_Wang')\nsplit = data.get_split(method='scaffold', seed=42, frac=[0.7, 0.1, 0.2])\ntrain, valid, test = split['train'], split['valid'], split['test']\nprint(f\"Train: {len(train)}, Valid: {len(valid)}, Test: {len(test)}\")\n# Train: ~640, Valid: ~91, Test: ~182\n\n# Evaluate predictions\nevaluator = Evaluator(name='MAE')\n# score = evaluator(test['Y'].values, predictions)\n```\n\n## Core API\n\n### Module 1: Single-Instance Prediction — Dataset Access\n\nLoad datasets for predicting properties of individual molecules or proteins.\n\n```python\nfrom tdc.single_pred import ADME, Tox, HTS, QM\n\n# ADME — pharmacokinetic properties\ndata = ADME(name='Caco2_Wang')       # Intestinal permeability (regression)\ndata = ADME(name='BBB_Martins')       # Blood-brain barrier (binary)\ndata = ADME(name='Lipophilicity_AstraZeneca')  # LogD (regression)\ndata = ADME(name='Solubility_AqSolDB')         # Aqueous solubility\n\n# Toxicity — adverse effects\ndata = Tox(name='hERG')              # Cardiotoxicity (binary)\ndata = Tox(name='AMES')              # Mutagenicity (binary)\ndata = Tox(name='DILI')              # Drug-induced liver injury\ndata = Tox(name='ClinTox')           # Clinical trial toxicity\n\n# Access data as DataFrame\ndf = data.get_data(format='df')\nprint(df.columns.tolist())\n# ['Drug_ID', 'Drug', 'Y'] — Drug is SMILES, Y is target label\nprint(f\"Dataset size: {len(df)}, Label range: [{df['Y'].min():.2f}, {df['Y'].max():.2f}]\")\n```\n\nOther single-prediction tasks: `HTS` (screening), `QM` (quantum mechanics), `Yields`, `Epitope`, `Develop`, `CRISPROutcome`.\n\n### Module 2: Multi-Instance Prediction — Interaction Datasets\n\nLoad datasets for predicting interactions between pairs of biomedical entities.\n\n```python\nfrom tdc.multi_pred import DTI, DDI, PPI\n\n# Drug-Target Interaction — binding affinity\ndata = DTI(name='BindingDB_Kd')      # 52,284 pairs, Kd values\ndata = DTI(name='DAVIS')             # 30,056 pairs, kinase binding\ndata = DTI(name='KIBA')              # 118,254 pairs, kinase bioactivity\n\n# Drug-Drug Interaction — interaction type prediction\ndata = DDI(name='DrugBank')           # 191,808 pairs, 86 interaction types\n\n# Protein-Protein Interaction\ndata = PPI(name='HuRI')\n\n# Multi-instance data format\ndf = data.get_data(format='df')\nprint(df.columns.tolist())\n# ['Drug_ID', 'Drug', 'Target_ID', 'Target', 'Y']\n# Drug=SMILES, Target=protein sequence, Y=binding affinity or class\n```\n\nOther multi-instance tasks: `GDA`, `DrugRes`, `DrugSyn`, `PeptideMHC`, `AntibodyAff`, `MTI`, `Catalyst`, `TrialOutcome`.\n\n### Module 3: Generation Tasks — Molecular Design\n\nLoad training sets and oracles for molecule generation and retrosynthesis.\n\n```python\nfrom tdc.generation import MolGen, RetroSyn, PairMolGen\nfrom tdc import Oracle\n\n# Molecule generation — training data\ndata = MolGen(name='ChEMBL_V29')     # 1.6M drug-like SMILES\nsplit = data.get_split()\ntrain_smiles = split['train']['Drug'].tolist()\n\n# Oracle scoring — evaluate generated molecules\noracle = Oracle(name='GSK3B')         # GSK3B inhibition predictor (0-1)\nscore = oracle('CC(C)Cc1ccc(cc1)C(C)C(O)=O')\nprint(f\"GSK3B score: {score:.4f}\")\n\n# Batch evaluation\nscores = oracle(['CCO', 'c1ccccc1', 'CC(=O)O'])\nprint(f\"Batch scores: {scores}\")\n\n# Retrosynthesis — reaction prediction\ndata = RetroSyn(name='USPTO')         # 1.9M reactions\nsplit = data.get_split()\n\n# Paired generation — prodrug design\ndata = PairMolGen(name='Prodrug')\n```\n\n### Module 4: Data Splits and Evaluation\n\nApply meaningful data splits and standardized evaluation metrics.\n\n```python\nfrom tdc.single_pred import ADME\nfrom tdc.multi_pred import DTI\nfrom tdc import Evaluator\n\n# Scaffold split — ensures chemical diversity between sets\ndata = ADME(name='Caco2_Wang')\nsplit = data.get_split(method='scaffold', seed=42, frac=[0.7, 0.1, 0.2])\n\n# Cold splits — for DTI (unseen drugs\u002Ftargets in test set)\ndata = DTI(name='BindingDB_Kd')\ncold_drug = data.get_split(method='cold_drug', seed=1)\ncold_target = data.get_split(method='cold_target', seed=1)\n\n# Verify no overlap in cold split\ntrain_drugs = set(cold_drug['train']['Drug_ID'])\ntest_drugs = set(cold_drug['test']['Drug_ID'])\nprint(f\"Drug overlap: {len(train_drugs & test_drugs)}\")  # 0\n\n# Evaluation metrics\neval_mae = Evaluator(name='MAE')\neval_auc = Evaluator(name='ROC-AUC')\neval_spearman = Evaluator(name='Spearman')\n# score = eval_mae(y_true, y_pred)\n```\n\n**Available split methods**: `random`, `scaffold` (Bemis-Murcko), `cold_drug`, `cold_target`, `cold_drug_target`, `temporal`.\n\n**Available metrics**: Classification — `ROC-AUC`, `PR-AUC`, `F1`, `Accuracy`, `Kappa`. Regression — `RMSE`, `MAE`, `R2`, `MSE`. Ranking — `Spearman`, `Pearson`. Multi-label — `Micro-F1`, `Macro-F1`.\n\n### Module 5: Benchmark Groups\n\nRun standardized multi-seed evaluation protocols for model comparison.\n\n```python\nfrom tdc.benchmark_group import admet_group\n\n# Load ADMET benchmark (22 datasets)\ngroup = admet_group(path='data\u002F')\n\n# Standard 5-seed evaluation protocol\nbenchmark = group.get('Caco2_Wang')\npredictions = {}\n\nfor seed in [1, 2, 3, 4, 5]:\n    train_df = benchmark['train']\n    valid_df = benchmark['valid']\n    test_df = benchmark['test']\n    # Train your model on train_df, tune on valid_df\n    # predictions[seed] = model.predict(test_df['Drug'])\n    predictions[seed] = test_df['Y'].values  # placeholder\n\n# Get benchmark results\nresults = group.evaluate(predictions)\nprint(f\"Mean MAE: {results['Caco2_Wang'][0]:.4f} ± {results['Caco2_Wang'][1]:.4f}\")\n```\n\n## Key Concepts\n\n### Dataset Organization\n\n| Category | Import Path | Task Examples | Data Format |\n|----------|-------------|---------------|-------------|\n| Single-Instance | `tdc.single_pred` | ADME, Tox, HTS, QM | Drug (SMILES) + Y (label) |\n| Multi-Instance | `tdc.multi_pred` | DTI, DDI, PPI, DrugSyn | Drug + Target + Y |\n| Generation | `tdc.generation` | MolGen, RetroSyn | SMILES collections |\n| Benchmark | `tdc.benchmark_group` | admet_group | Curated splits |\n\n### Oracle Categories\n\n| Category | Examples | Speed | Output Range |\n|----------|----------|-------|-------------|\n| Biochemical | DRD2, GSK3B, JNK3, 5HT2A | Medium (ML) | 0-1 probability |\n| Physicochemical | QED, SA, LogP, MW | Fast (rule-based) | Varies by metric |\n| Composite | Isomer_Meta, Median1\u002F2, Rediscovery | Medium | 0-1 combined |\n| Specialized | ASKCOS, Docking, Vina | Slow (external) | Varies |\n\n### Data Processing Utilities\n\n| Utility | Function | Example |\n|---------|----------|---------|\n| Format conversion | `MolConvert(src, dst)` | SMILES → PyG, ECFP, SELFIES, DGL |\n| Molecule filters | `MolFilter(filters)` | PAINS, BMS, Glaxo, drug-likeness |\n| Label binarization | `label_transform()` | Continuous → binary at threshold |\n| Unit conversion | `label_transform(from_unit, to_unit)` | nM → pIC50 |\n| ID resolution | `cid2smiles()`, `uniprot2seq()` | PubChem CID → SMILES |\n| Dataset listing | `retrieve_dataset_names(task)` | List all ADME datasets |\n\n## Common Workflows\n\n### Workflow 1: Multi-Seed ADME Model Evaluation\n\n```python\nfrom tdc.single_pred import ADME\nfrom tdc import Evaluator\nimport numpy as np\n\ndata = ADME(name='Caco2_Wang')\nevaluator = Evaluator(name='MAE')\n\nresults = []\nfor seed in [1, 2, 3, 4, 5]:\n    split = data.get_split(method='scaffold', seed=seed)\n    train, valid, test = split['train'], split['valid'], split['test']\n    # model.fit(train['Drug'], train['Y'])\n    # preds = model.predict(test['Drug'])\n    preds = test['Y'].values + np.random.normal(0, 0.1, len(test))  # placeholder\n    score = evaluator(test['Y'].values, preds)\n    results.append(score)\n    print(f\"Seed {seed}: MAE = {score:.4f}\")\n\nprint(f\"Mean MAE: {np.mean(results):.4f} ± {np.std(results):.4f}\")\n```\n\n### Workflow 2: Multi-Objective Molecular Scoring\n\n```python\nfrom tdc import Oracle\nimport numpy as np\n\n# Define multi-objective scoring\noracles = {\n    'QED': (Oracle(name='QED'), 0.3),      # drug-likeness\n    'SA': (Oracle(name='SA'), 0.3),         # synthetic accessibility\n    'GSK3B': (Oracle(name='GSK3B'), 0.4),   # target activity\n}\n\ntest_smiles = ['CC(C)Cc1ccc(cc1)C(C)C(O)=O', 'c1ccc2c(c1)cc1ccc3cccc4ccc2c1c34']\n\nfor smi in test_smiles:\n    scores = {}\n    weighted_sum = 0\n    for name, (oracle, weight) in oracles.items():\n        score = oracle(smi)\n        scores[name] = score\n        weighted_sum += score * weight\n    print(f\"SMILES: {smi[:30]}...\")\n    print(f\"  Scores: {scores}\")\n    print(f\"  Weighted: {weighted_sum:.4f}\")\n```\n\n### Workflow 3: Cold-Split DTI Evaluation (text-only)\n\n1. Load DTI dataset — `DTI(name='BindingDB_Kd')` (Core API Module 2)\n2. Apply cold_drug split — `data.get_split(method='cold_drug', seed=seed)` (Core API Module 4)\n3. Verify zero drug overlap between train and test sets (Module 4 overlap check)\n4. Train model on train set, predict on test set\n5. Evaluate with Spearman correlation — `Evaluator(name='Spearman')` (Module 4)\n6. Repeat for 5 seeds and report mean ± std (same pattern as Workflow 1)\n\n## Key Parameters\n\n| Parameter | Function\u002FModule | Default | Range\u002FOptions | Effect |\n|-----------|----------------|---------|---------------|--------|\n| `method` | `get_split()` | `'scaffold'` | `random`, `scaffold`, `cold_drug`, `cold_target`, `temporal` | Split strategy for train\u002Ftest |\n| `seed` | `get_split()` | `42` | 1-5 for benchmarks | Reproducibility; use 5 seeds for benchmarks |\n| `frac` | `get_split()` | `[0.7, 0.1, 0.2]` | Sum must equal 1.0 | Train\u002Fvalid\u002Ftest proportions |\n| `name` | `Evaluator()` | — | `MAE`, `RMSE`, `ROC-AUC`, `Spearman`, etc. | Evaluation metric |\n| `name` | `Oracle()` | — | `QED`, `SA`, `GSK3B`, `DRD2`, etc. | Scoring function for molecules |\n| `src`\u002F`dst` | `MolConvert()` | — | `SMILES`, `SELFIES`, `PyG`, `DGL`, `ECFP4` | Molecular representation formats |\n| `path` | `admet_group()` | `'data\u002F'` | Any directory | Dataset download\u002Fcache location |\n| `format` | `get_data()` | `'df'` | `'df'`, `'dict'` | Output data format |\n\n## Best Practices\n\n1. **Always use scaffold splits** for molecular property prediction — random splits leak structural information and inflate performance metrics\n2. **Report 5-seed evaluations** with mean ± std — single-seed results are unreliable for method comparison\n3. **Use cold splits for interaction prediction** — `cold_drug` tests generalization to unseen drugs, `cold_target` to unseen targets\n4. **Filter molecules early** with `MolFilter` (PAINS, drug-likeness) before training to remove problematic compounds\n5. **Normalize oracles appropriately** — QED returns 0-1, SA returns 1-10 (lower is better), binding scores vary. Check oracle documentation before combining\n6. **Cache datasets** by specifying a persistent `path` — avoids re-downloading large datasets across sessions\n\n## Common Recipes\n\n### Recipe 1: Dataset Exploration and Statistics\n\n```python\nfrom tdc.single_pred import ADME\nfrom tdc.utils import retrieve_dataset_names\n\n# List all ADME datasets\ndatasets = retrieve_dataset_names('ADME')\nprint(f\"Available ADME datasets: {datasets}\")\n\n# Load and inspect\ndata = ADME(name='Caco2_Wang')\ndf = data.get_data(format='df')\nprint(f\"Size: {len(df)}\")\nprint(f\"Label stats: mean={df['Y'].mean():.2f}, std={df['Y'].std():.2f}\")\nprint(f\"SMILES example: {df['Drug'].iloc[0]}\")\n```\n\n### Recipe 2: Molecule Format Conversion Pipeline\n\n```python\nfrom tdc.chem_utils import MolConvert\n\n# SMILES to multiple representations\nsmiles = 'CC(C)Cc1ccc(cc1)C(C)C(O)=O'\n\nconverter_ecfp = MolConvert(src='SMILES', dst='ECFP4')\nconverter_selfies = MolConvert(src='SMILES', dst='SELFIES')\n\necfp = converter_ecfp(smiles)\nselfies = converter_selfies(smiles)\nprint(f\"ECFP4 shape: {ecfp.shape}\")    # (1024,) binary fingerprint\nprint(f\"SELFIES: {selfies}\")\n```\n\n### Recipe 3: Custom Oracle with Constraint Satisfaction\n\n```python\nfrom tdc import Oracle\n\n# Define property constraints\nconstraints = {\n    'QED': (Oracle(name='QED'), 0.5, 1.0),       # min, max\n    'SA': (Oracle(name='SA'), 1.0, 4.0),\n    'LogP': (Oracle(name='LogP'), -0.5, 5.0),\n}\n\ndef check_constraints(smiles):\n    \"\"\"Check if molecule satisfies all property constraints.\"\"\"\n    results = {}\n    all_pass = True\n    for name, (oracle, lo, hi) in constraints.items():\n        score = oracle(smiles)\n        passed = lo \u003C= score \u003C= hi\n        results[name] = {'score': score, 'passed': passed}\n        all_pass = all_pass and passed\n    return all_pass, results\n\npassed, details = check_constraints('CC(C)Cc1ccc(cc1)C(C)C(O)=O')\nfor name, info in details.items():\n    status = \"PASS\" if info['passed'] else \"FAIL\"\n    print(f\"{name}: {info['score']:.3f} [{status}]\")\n```\n\n## Troubleshooting\n\n| Problem | Cause | Solution |\n|---------|-------|----------|\n| `ModuleNotFoundError: tdc` | Package not installed | `uv pip install PyTDC` |\n| Scaffold split fails | Missing RDKit dependency | `uv pip install rdkit` for scaffold decomposition |\n| Dataset download timeout | Large dataset or slow connection | Set `path='data\u002F'` for persistent cache; retry |\n| `KeyError` on dataset name | Wrong name or task category | Use `retrieve_dataset_names('ADME')` to list valid names |\n| Oracle returns NaN | Invalid SMILES or RDKit parse failure | Validate SMILES with RDKit `MolFromSmiles()` first |\n| Cold split empty test set | Too few unique entities | Use `frac=[0.7, 0.1, 0.2]` with larger datasets |\n| Benchmark evaluation error | Wrong prediction format | Pass `dict` with seeds as keys: `{1: preds1, 2: preds2, ...}` |\n| Memory error on large dataset | Full dataset loaded to memory | Process in chunks or use smaller split fractions |\n| PyG conversion fails | torch-geometric not installed | `uv pip install torch-geometric` for graph conversion |\n\n## Bundled Resources\n\n### references\u002Fdatasets_catalog.md\nCovers: complete catalog of all TDC datasets organized by task category (single-instance, multi-instance, generation) with dataset names, sizes, label types, and data sources. Relocated inline: top ADME\u002FTox\u002FDTI datasets with code examples consolidated into Core API Modules 1-2. Omitted: None — all dataset entries preserved in catalog format.\n\n### references\u002Foracles_utilities.md\nCovers: detailed oracle documentation (all 17+ oracles with parameters, speed tiers, output ranges, custom oracle template) and data processing utilities (format conversion targets, molecule filter types, label transformation, entity resolution). Consolidated from original `oracles.md` + `utilities.md`. Relocated inline: Quick Start oracle usage pattern, top evaluation metrics, split methods, MolConvert pattern → Core API Modules 3-4. Omitted: distribution learning KS-test example — niche statistical comparison; leaderboard submission guide — platform-specific.\n\n### Script Disposition\n- `load_and_split_data.py` (215 lines): scaffold\u002Fcold split patterns → Core API Module 4; custom split fractions → Key Parameters; evaluation examples → Workflow 1. Thin wrappers around `get_split()` and `Evaluator()`.\n- `benchmark_evaluation.py` (328 lines): 5-seed protocol → Core API Module 5 + Workflow 1; multi-dataset evaluation → Workflow 1 pattern; leaderboard guide → omitted (platform-specific).\n- `molecular_generation.py` (405 lines): single\u002Fbatch oracle usage → Core API Module 3; multi-objective scoring → Workflow 2; constraint satisfaction → Recipe 3; distribution learning → omitted (niche).\n\n## Related Skills\n\n- **chembl-database-bioactivity** — for querying ChEMBL compound\u002Ftarget\u002Factivity data directly\n- **molfeat** — for advanced molecular featurization beyond TDC's built-in MolConvert\n- **rdkit-cheminformatics** — for molecular manipulation, substructure search, descriptor calculation\n\n## References\n\n- TDC Official Website: https:\u002F\u002Ftdcommons.ai\n- TDC Documentation: https:\u002F\u002Ftdc.readthedocs.io\n- GitHub Repository: https:\u002F\u002Fgithub.com\u002Fmims-harvard\u002FTDC\n- Huang et al. \"Therapeutics Data Commons\" NeurIPS 2021 Datasets and Benchmarks\n","# PyTDC (Therapeutics Data Commons)\n\n## Overview\n\nPyTDC is an open-science platform providing AI-ready datasets and benchmarks for drug discovery. It organizes therapeutics data into three categories: single-instance prediction (molecular\u002Fprotein properties), multi-instance prediction (drug-target interactions), and generation (molecule design, retrosynthesis). All datasets come with standardized splits, evaluation metrics, and molecular oracles.\n\n## When to Use\n\n- Loading curated ADME, toxicity, or bioactivity datasets for ML model training\n- Benchmarking drug discovery models with standardized 5-seed evaluation protocols\n- Predicting drug-target or drug-drug interactions with proper cold-split evaluation\n- Generating novel molecules and scoring them with molecular oracles (QED, SA, DRD2, GSK3B)\n- Accessing scaffold-based or temporal train\u002Ftest splits for pharmaceutical ML\n- Converting molecular representations (SMILES to PyG graphs, ECFP fingerprints, SELFIES)\n- For chemical database queries (compound search, bioactivity), use `chembl-database-bioactivity` instead\n- For molecular featurization beyond format conversion, use `molfeat` instead\n\n## Prerequisites\n\n```bash\nuv pip install PyTDC\n# Core deps: numpy, pandas, scikit-learn, tqdm, fuzzywuzzy\n# Optional: rdkit (scaffold splits), torch-geometric (PyG conversion)\n```\n\n**API Note**: TDC downloads datasets on first access (~10-500 MB per dataset). Specify `path='data\u002F'` to control download location. No API key required.\n\n## Quick Start\n\n```python\nfrom tdc.single_pred import ADME\nfrom tdc import Evaluator\n\n# Load dataset with scaffold split\ndata = ADME(name='Caco2_Wang')\nsplit = data.get_split(method='scaffold', seed=42, frac=[0.7, 0.1, 0.2])\ntrain, valid, test = split['train'], split['valid'], split['test']\nprint(f\"Train: {len(train)}, Valid: {len(valid)}, Test: {len(test)}\")\n# Train: ~640, Valid: ~91, Test: ~182\n\n# Evaluate predictions\nevaluator = Evaluator(name='MAE')\n# score = evaluator(test['Y'].values, predictions)\n```\n\n## Core API\n\n### Module 1: Single-Instance Prediction — Dataset Access\n\nLoad datasets for predicting properties of individual molecules or proteins.\n\n```python\nfrom tdc.single_pred import ADME, Tox, HTS, QM\n\n# ADME — pharmacokinetic properties\ndata = ADME(name='Caco2_Wang')       # Intestinal permeability (regression)\ndata = ADME(name='BBB_Martins')       # Blood-brain barrier (binary)\ndata = ADME(name='Lipophilicity_AstraZeneca')  # LogD (regression)\ndata = ADME(name='Solubility_AqSolDB')         # Aqueous solubility\n\n# Toxicity — adverse effects\ndata = Tox(name='hERG')              # Cardiotoxicity (binary)\ndata = Tox(name='AMES')              # Mutagenicity (binary)\ndata = Tox(name='DILI')              # Drug-induced liver injury\ndata = Tox(name='ClinTox')           # Clinical trial toxicity\n\n# Access data as DataFrame\ndf = data.get_data(format='df')\nprint(df.columns.tolist())\n# ['Drug_ID', 'Drug', 'Y'] — Drug is SMILES, Y is target label\nprint(f\"Dataset size: {len(df)}, Label range: [{df['Y'].min():.2f}, {df['Y'].max():.2f}]\")\n```\n\nOther single-prediction tasks: `HTS` (screening), `QM` (quantum mechanics), `Yields`, `Epitope`, `Develop`, `CRISPROutcome`.\n\n### Module 2: Multi-Instance Prediction — Interaction Datasets\n\nLoad datasets for predicting interactions between pairs of biomedical entities.\n\n```python\nfrom tdc.multi_pred import DTI, DDI, PPI\n\n# Drug-Target Interaction — binding affinity\ndata = DTI(name='BindingDB_Kd')      # 52,284 pairs, Kd values\ndata = DTI(name='DAVIS')             # 30,056 pairs, kinase binding\ndata = DTI(name='KIBA')              # 118,254 pairs, kinase bioactivity\n\n# Drug-Drug Interaction — interaction type prediction\ndata = DDI(name='DrugBank')           # 191,808 pairs, 86 interaction types\n\n# Protein-Protein Interaction\ndata = PPI(name='HuRI')\n\n# Multi-instance data format\ndf = data.get_data(format='df')\nprint(df.columns.tolist())\n# ['Drug_ID', 'Drug', 'Target_ID', 'Target', 'Y']\n# Drug=SMILES, Target=protein sequence, Y=binding affinity or class\n```\n\nOther multi-instance tasks: `GDA`, `DrugRes`, `DrugSyn`, `PeptideMHC`, `AntibodyAff`, `MTI`, `Catalyst`, `TrialOutcome`.\n\n### Module 3: Generation Tasks — Molecular Design\n\nLoad training sets and oracles for molecule generation and retrosynthesis.\n\n```python\nfrom tdc.generation import MolGen, RetroSyn, PairMolGen\nfrom tdc import Oracle\n\n# Molecule generation — training data\ndata = MolGen(name='ChEMBL_V29')     # 1.6M drug-like SMILES\nsplit = data.get_split()\ntrain_smiles = split['train']['Drug'].tolist()\n\n# Oracle scoring — evaluate generated molecules\noracle = Oracle(name='GSK3B')         # GSK3B inhibition predictor (0-1)\nscore = oracle('CC(C)Cc1ccc(cc1)C(C)C(O)=O')\nprint(f\"GSK3B score: {score:.4f}\")\n\n# Batch evaluation\nscores = oracle(['CCO', 'c1ccccc1', 'CC(=O)O'])\nprint(f\"Batch scores: {scores}\")\n\n# Retrosynthesis — reaction prediction\ndata = RetroSyn(name='USPTO')         # 1.9M reactions\nsplit = data.get_split()\n\n# Paired generation — prodrug design\ndata = PairMolGen(name='Prodrug')\n```\n\n### Module 4: Data Splits and Evaluation\n\nApply meaningful data splits and standardized evaluation metrics.\n\n```python\nfrom tdc.single_pred import ADME\nfrom tdc.multi_pred import DTI\nfrom tdc import Evaluator\n\n# Scaffold split — ensures chemical diversity between sets\ndata = ADME(name='Caco2_Wang')\nsplit = data.get_split(method='scaffold', seed=42, frac=[0.7, 0.1, 0.2])\n\n# Cold splits — for DTI (unseen drugs\u002Ftargets in test set)\ndata = DTI(name='BindingDB_Kd')\ncold_drug = data.get_split(method='cold_drug', seed=1)\ncold_target = data.get_split(method='cold_target', seed=1)\n\n# Verify no overlap in cold split\ntrain_drugs = set(cold_drug['train']['Drug_ID'])\ntest_drugs = set(cold_drug['test']['Drug_ID'])\nprint(f\"Drug overlap: {len(train_drugs & test_drugs)}\")  # 0\n\n# Evaluation metrics\neval_mae = Evaluator(name='MAE')\neval_auc = Evaluator(name='ROC-AUC')\neval_spearman = Evaluator(name='Spearman')\n# score = eval_mae(y_true, y_pred)\n```\n\n**Available split methods**: `random`, `scaffold` (Bemis-Murcko), `cold_drug`, `cold_target`, `cold_drug_target`, `temporal`.\n\n**Available metrics**: Classification — `ROC-AUC`, `PR-AUC`, `F1`, `Accuracy`, `Kappa`. Regression — `RMSE`, `MAE`, `R2`, `MSE`. Ranking — `Spearman`, `Pearson`. Multi-label — `Micro-F1`, `Macro-F1`.\n\n### Module 5: Benchmark Groups\n\nRun standardized multi-seed evaluation protocols for model comparison.\n\n```python\nfrom tdc.benchmark_group import admet_group\n\n# Load ADMET benchmark (22 datasets)\ngroup = admet_group(path='data\u002F')\n\n# Standard 5-seed evaluation protocol\nbenchmark = group.get('Caco2_Wang')\npredictions = {}\n\nfor seed in [1, 2, 3, 4, 5]:\n    train_df = benchmark['train']\n    valid_df = benchmark['valid']\n    test_df = benchmark['test']\n    # Train your model on train_df, tune on valid_df\n    # predictions[seed] = model.predict(test_df['Drug'])\n    predictions[seed] = test_df['Y'].values  # placeholder\n\n# Get benchmark results\nresults = group.evaluate(predictions)\nprint(f\"Mean MAE: {results['Caco2_Wang'][0]:.4f} ± {results['Caco2_Wang'][1]:.4f}\")\n```\n\n## Key Concepts\n\n### Dataset Organization\n\n| Category | Import Path | Task Examples | Data Format |\n|----------|-------------|---------------|-------------|\n| Single-Instance | `tdc.single_pred` | ADME, Tox, HTS, QM | Drug (SMILES) + Y (label) |\n| Multi-Instance | `tdc.multi_pred` | DTI, DDI, PPI, DrugSyn | Drug + Target + Y |\n| Generation | `tdc.generation` | MolGen, RetroSyn | SMILES collections |\n| Benchmark | `tdc.benchmark_group` | admet_group | Curated splits |\n\n### Oracle Categories\n\n| Category | Examples | Speed | Output Range |\n|----------|----------|-------|-------------|\n| Biochemical | DRD2, GSK3B, JNK3, 5HT2A | Medium (ML) | 0-1 probability |\n| Physicochemical | QED, SA, LogP, MW | Fast (rule-based) | Varies by metric |\n| Composite | Isomer_Meta, Median1\u002F2, Rediscovery | Medium | 0-1 combined |\n| Specialized | ASKCOS, Docking, Vina | Slow (external) | Varies |\n\n### Data Processing Utilities\n\n| Utility | Function | Example |\n|---------|----------|---------|\n| Format conversion | `MolConvert(src, dst)` | SMILES → PyG, ECFP, SELFIES, DGL |\n| Molecule filters | `MolFilter(filters)` | PAINS, BMS, Glaxo, drug-likeness |\n| Label binarization | `label_transform()` | Continuous → binary at threshold |\n| Unit conversion | `label_transform(from_unit, to_unit)` | nM → pIC50 |\n| ID resolution | `cid2smiles()`, `uniprot2seq()` | PubChem CID → SMILES |\n| Dataset listing | `retrieve_dataset_names(task)` | List all ADME datasets |\n\n## Common Workflows\n\n### Workflow 1: Multi-Seed ADME Model Evaluation\n\n```python\nfrom tdc.single_pred import ADME\nfrom tdc import Evaluator\nimport numpy as np\n\ndata = ADME(name='Caco2_Wang')\nevaluator = Evaluator(name='MAE')\n\nresults = []\nfor seed in [1, 2, 3, 4, 5]:\n    split = data.get_split(method='scaffold', seed=seed)\n    train, valid, test = split['train'], split['valid'], split['test']\n    # model.fit(train['Drug'], train['Y'])\n    # preds = model.predict(test['Drug'])\n    preds = test['Y'].values + np.random.normal(0, 0.1, len(test))  # placeholder\n    score = evaluator(test['Y'].values, preds)\n    results.append(score)\n    print(f\"Seed {seed}: MAE = {score:.4f}\")\n\nprint(f\"Mean MAE: {np.mean(results):.4f} ± {np.std(results):.4f}\")\n```\n\n### Workflow 2: Multi-Objective Molecular Scoring\n\n```python\nfrom tdc import Oracle\nimport numpy as np\n\n# Define multi-objective scoring\noracles = {\n    'QED': (Oracle(name='QED'), 0.3),      # drug-likeness\n    'SA': (Oracle(name='SA'), 0.3),         # synthetic accessibility\n    'GSK3B': (Oracle(name='GSK3B'), 0.4),   # target activity\n}\n\ntest_smiles = ['CC(C)Cc1ccc(cc1)C(C)C(O)=O', 'c1ccc2c(c1)cc1ccc3cccc4ccc2c1c34']\n\nfor smi in test_smiles:\n    scores = {}\n    weighted_sum = 0\n    for name, (oracle, weight) in oracles.items():\n        score = oracle(smi)\n        scores[name] = score\n        weighted_sum += score * weight\n    print(f\"SMILES: {smi[:30]}...\")\n    print(f\"  Scores: {scores}\")\n    print(f\"  Weighted: {weighted_sum:.4f}\")\n```\n\n### Workflow 3: Cold-Split DTI Evaluation (text-only)\n\n1. Load DTI dataset — `DTI(name='BindingDB_Kd')` (Core API Module 2)\n2. Apply cold_drug split — `data.get_split(method='cold_drug', seed=seed)` (Core API Module 4)\n3. Verify zero drug overlap between train and test sets (Module 4 overlap check)\n4. Train model on train set, predict on test set\n5. Evaluate with Spearman correlation — `Evaluator(name='Spearman')` (Module 4)\n6. Repeat for 5 seeds and report mean ± std (same pattern as Workflow 1)\n\n## Key Parameters\n\n| Parameter | Function\u002FModule | Default | Range\u002FOptions | Effect |\n|-----------|----------------|---------|---------------|--------|\n| `method` | `get_split()` | `'scaffold'` | `random`, `scaffold`, `cold_drug`, `cold_target`, `temporal` | Split strategy for train\u002Ftest |\n| `seed` | `get_split()` | `42` | 1-5 for benchmarks | Reproducibility; use 5 seeds for benchmarks |\n| `frac` | `get_split()` | `[0.7, 0.1, 0.2]` | Sum must equal 1.0 | Train\u002Fvalid\u002Ftest proportions |\n| `name` | `Evaluator()` | — | `MAE`, `RMSE`, `ROC-AUC`, `Spearman`, etc. | Evaluation metric |\n| `name` | `Oracle()` | — | `QED`, `SA`, `GSK3B`, `DRD2`, etc. | Scoring function for molecules |\n| `src`\u002F`dst` | `MolConvert()` | — | `SMILES`, `SELFIES`, `PyG`, `DGL`, `ECFP4` | Molecular representation formats |\n| `path` | `admet_group()` | `'data\u002F'` | Any directory | Dataset download\u002Fcache location |\n| `format` | `get_data()` | `'df'` | `'df'`, `'dict'` | Output data format |\n\n## Best Practices\n\n1. **Always use scaffold splits** for molecular property prediction — random splits leak structural information and inflate performance metrics\n2. **Report 5-seed evaluations** with mean ± std — single-seed results are unreliable for method comparison\n3. **Use cold splits for interaction prediction** — `cold_drug` tests generalization to unseen drugs, `cold_target` to unseen targets\n4. **Filter molecules early** with `MolFilter` (PAINS, drug-likeness) before training to remove problematic compounds\n5. **Normalize oracles appropriately** — QED returns 0-1, SA returns 1-10 (lower is better), binding scores vary. Check oracle documentation before combining\n6. **Cache datasets** by specifying a persistent `path` — avoids re-downloading large datasets across sessions\n\n## Common Recipes\n\n### Recipe 1: Dataset Exploration and Statistics\n\n```python\nfrom tdc.single_pred import ADME\nfrom tdc.utils import retrieve_dataset_names\n\n# List all ADME datasets\ndatasets = retrieve_dataset_names('ADME')\nprint(f\"Available ADME datasets: {datasets}\")\n\n# Load and inspect\ndata = ADME(name='Caco2_Wang')\ndf = data.get_data(format='df')\nprint(f\"Size: {len(df)}\")\nprint(f\"Label stats: mean={df['Y'].mean():.2f}, std={df['Y'].std():.2f}\")\nprint(f\"SMILES example: {df['Drug'].iloc[0]}\")\n```\n\n### Recipe 2: Molecule Format Conversion Pipeline\n\n```python\nfrom tdc.chem_utils import MolConvert\n\n# SMILES to multiple representations\nsmiles = 'CC(C)Cc1ccc(cc1)C(C)C(O)=O'\n\nconverter_ecfp = MolConvert(src='SMILES', dst='ECFP4')\nconverter_selfies = MolConvert(src='SMILES', dst='SELFIES')\n\necfp = converter_ecfp(smiles)\nselfies = converter_selfies(smiles)\nprint(f\"ECFP4 shape: {ecfp.shape}\")    # (1024,) binary fingerprint\nprint(f\"SELFIES: {selfies}\")\n```\n\n### Recipe 3: Custom Oracle with Constraint Satisfaction\n\n```python\nfrom tdc import Oracle\n\n# Define property constraints\nconstraints = {\n    'QED': (Oracle(name='QED'), 0.5, 1.0),       # min, max\n    'SA': (Oracle(name='SA'), 1.0, 4.0),\n    'LogP': (Oracle(name='LogP'), -0.5, 5.0),\n}\n\ndef check_constraints(smiles):\n    \"\"\"Check if molecule satisfies all property constraints.\"\"\"\n    results = {}\n    all_pass = True\n    for name, (oracle, lo, hi) in constraints.items():\n        score = oracle(smiles)\n        passed = lo \u003C= score \u003C= hi\n        results[name] = {'score': score, 'passed': passed}\n        all_pass = all_pass and passed\n    return all_pass, results\n\npassed, details = check_constraints('CC(C)Cc1ccc(cc1)C(C)C(O)=O')\nfor name, info in details.items():\n    status = \"PASS\" if info['passed'] else \"FAIL\"\n    print(f\"{name}: {info['score']:.3f} [{status}]\")\n```\n\n## Troubleshooting\n\n| Problem | Cause | Solution |\n|---------|-------|----------|\n| `ModuleNotFoundError: tdc` | Package not installed | `uv pip install PyTDC` |\n| Scaffold split fails | Missing RDKit dependency | `uv pip install rdkit` for scaffold decomposition |\n| Dataset download timeout | Large dataset or slow connection | Set `path='data\u002F'` for persistent cache; retry |\n| `KeyError` on dataset name | Wrong name or task category | Use `retrieve_dataset_names('ADME')` to list valid names |\n| Oracle returns NaN | Invalid SMILES or RDKit parse failure | Validate SMILES with RDKit `MolFromSmiles()` first |\n| Cold split empty test set | Too few unique entities | Use `frac=[0.7, 0.1, 0.2]` with larger datasets |\n| Benchmark evaluation error | Wrong prediction format | Pass `dict` with seeds as keys: `{1: preds1, 2: preds2, ...}` |\n| Memory error on large dataset | Full dataset loaded to memory | Process in chunks or use smaller split fractions |\n| PyG conversion fails | torch-geometric not installed | `uv pip install torch-geometric` for graph conversion |\n\n## Bundled Resources\n\n### references\u002Fdatasets_catalog.md\nCovers: complete catalog of all TDC datasets organized by task category (single-instance, multi-instance, generation) with dataset names, sizes, label types, and data sources. Relocated inline: top ADME\u002FTox\u002FDTI datasets with code examples consolidated into Core API Modules 1-2. Omitted: None — all dataset entries preserved in catalog format.\n\n### references\u002Foracles_utilities.md\nCovers: detailed oracle documentation (all 17+ oracles with parameters, speed tiers, output ranges, custom oracle template) and data processing utilities (format conversion targets, molecule filter types, label transformation, entity resolution). Consolidated from original `oracles.md` + `utilities.md`. Relocated inline: Quick Start oracle usage pattern, top evaluation metrics, split methods, MolConvert pattern → Core API Modules 3-4. Omitted: distribution learning KS-test example — niche statistical comparison; leaderboard submission guide — platform-specific.\n\n### Script Disposition\n- `load_and_split_data.py` (215 lines): scaffold\u002Fcold split patterns → Core API Module 4; custom split fractions → Key Parameters; evaluation examples → Workflow 1. Thin wrappers around `get_split()` and `Evaluator()`.\n- `benchmark_evaluation.py` (328 lines): 5-seed protocol → Core API Module 5 + Workflow 1; multi-dataset evaluation → Workflow 1 pattern; leaderboard guide → omitted (platform-specific).\n- `molecular_generation.py` (405 lines): single\u002Fbatch oracle usage → Core API Module 3; multi-objective scoring → Workflow 2; constraint satisfaction → Recipe 3; distribution learning → omitted (niche).\n\n## Related Skills\n\n- **chembl-database-bioactivity** — for querying ChEMBL compound\u002Ftarget\u002Factivity data directly\n- **molfeat** — for advanced molecular featurization beyond TDC's built-in MolConvert\n- **rdkit-cheminformatics** — for molecular manipulation, substructure search, descriptor calculation\n\n## References\n\n- TDC Official Website: https:\u002F\u002Ftdcommons.ai\n- TDC Documentation: https:\u002F\u002Ftdc.readthedocs.io\n- GitHub Repository: https:\u002F\u002Fgithub.com\u002Fmims-harvard\u002FTDC\n- Huang et al. \"Therapeutics Data Commons\" NeurIPS 2021 Datasets and Benchmarks\n",{"name":5,"license":15,"description":7},"master","63f17866001c5ca854bf2fa0b678ce277133de5b","skills\u002Fawesome-skills\u002Fpytdc-therapeutics-data-commons",0,"https:\u002F\u002Fgithub.com\u002Fbg-szy\u002FTOP-SKILLS","https:\u002F\u002Fbg-szy.github.io\u002FTOP-SKILLS\u002F","https:\u002F\u002Fgithub.com\u002Fbg-szy","https:\u002F\u002Fwww.ecust.edu.cn\u002F","a student of East China University of Science and Technology","2026-11-15 03:38:02",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add bg-szy\u002FTOP-SKILLS --skill pytdc-therapeutics-data-commons",{"tool":40,"command":41},"bunx","bunx skills add bg-szy\u002FTOP-SKILLS --skill pytdc-therapeutics-data-commons",{"tool":43,"command":44},"pnpm","pnpm dlx skills add bg-szy\u002FTOP-SKILLS --skill pytdc-therapeutics-data-commons","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fbg-szy\u002FTOP-SKILLS\u002Fblob\u002Fmaster\u002Fskills\u002Fawesome-skills\u002Fpytdc-therapeutics-data-commons\u002FSKILL.md",true,[],[49,57,64,71,78,85],{"skillNo":50,"name":51,"nameZh":6,"description":52,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":56},"01M7DXW9N826PKXK174PVW3038","serpmaven","Scrape Google search results (SERP) to JSON with SERP Maven (serpmaven) - a real Chrome browser on a residential proxy with sticky sessions. Run it as a CLI, a Python library, a local HTTP search API (`serve`), an MCP server or a Docker image; set the proxy login, workers, pages and cost tracking. Use when the user wants Google results, rank tracking, a self-hosted SERP API, or live Google search for an AI agent.","https:\u002F\u002Fgithub.com\u002Fbg-szy\u002FTOP-SKILLS\u002Fblob\u002Fmaster\u002Fskills\u002Fmarketplace\u002Fserpmaven\u002FSKILL.md",[],[],"2026-11-15 14:19:17",{"skillNo":58,"name":59,"nameZh":6,"description":60,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":61,"categories":62,"keywords":63,"publishedAt":56},"01M7DXVHSTWDV65475QEQ915M7","revnu-landing-page-audit","Find out why a landing page isn't turning visitors into signups, demos or sales. Audits one page against what the ad, post or search that sends people there promised, then checks the first screen, proof, call to action, friction and mobile. Returns at most 7 prioritized fixes, plus what is fine and what could not be checked; rewrites specific lines and fixes the page in the repo when asked. Never promises a conversion lift. Works without a Revnu account. Use when the user says or asks: \"audit my landing page\", \"why isn't this page converting\", \"review my homepage\", \"my ads get clicks but no signups\", \"people bounce from the pricing page\", \"CRO review\", \"improve conversion rate\", \"is my hero section clear\", \"roast my landing page\", \"rewrite my headline\", \"what's wrong with this page\", \"landing page feedback\", \"make the signup page convert better\". Not for SEO or AI-search audits, page speed engineering on its own, or building a new page from scratch.","https:\u002F\u002Fgithub.com\u002Fbg-szy\u002FTOP-SKILLS\u002Fblob\u002Fmaster\u002Fskills\u002Fmarketplace\u002Frevnu-landing-page-audit\u002FSKILL.md",[],[],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":56},"01M7DXV8X2D7B3VTD4HHYYHTGR","revnu-growth-context","Write down once who buys, what they're stuck on, what it costs, what proof you have and what you've already tried, so every growth task starts from facts instead of a questionnaire. Creates or updates .agents\u002Frevnu-growth.md from the repo, the live site and the user, and imports an existing product-marketing.md. Use when the user says \"set up our marketing context\", \"who is our ideal customer\", \"write our ICP\", \"describe the product for marketing\", \"positioning doc\", \"what's our brand voice\", \"remember this about our customers\", \"log what we tried\", \"what have we learned from marketing\", \"we changed pricing, update the context\", \"update our growth notes\", \"summarize the business for the agent\", \"import my product-marketing.md\", \"start a growth doc\", or before a first marketing, SEO, outreach or launch task when no context file exists. Works without a Revnu account.","https:\u002F\u002Fgithub.com\u002Fbg-szy\u002FTOP-SKILLS\u002Fblob\u002Fmaster\u002Fskills\u002Fmarketplace\u002Frevnu-growth-context\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":56},"01M7DXV0DGJ1G0A60PAHC6J3C5","revnu-get-listed","Get onto the third-party pages AI answers and search cite when buyers ask for the best tool in your category: \"best X\" roundups, directories, review sites, GitHub awesome lists, community threads and comparison pages. Finds which ones are cited, ranks them by fit and effort, and drafts each submission, PR or pitch by the target's own rules. Use when the user says \"get us listed\", \"which directories should we submit to\", \"how do we get into best-of lists\", \"why does ChatGPT recommend competitors and not us\", \"get mentioned in AI answers\", \"submit to awesome lists\", \"get on G2\", \"where are our competitors listed\", \"pitch listicle authors\", \"find backlink opportunities\", \"launch directories\", \"get into roundups\", \"Product Hunt alternatives\", \"where should we be listed\", \"get cited by Perplexity\". Never submits, pays or posts without a yes. Works without a Revnu account.","https:\u002F\u002Fgithub.com\u002Fbg-szy\u002FTOP-SKILLS\u002Fblob\u002Fmaster\u002Fskills\u002Fmarketplace\u002Frevnu-get-listed\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":56},"01M7DXTSPMGT9D6DSS0HP9BSTA","revnu-ai-readiness-audit","Audit whether your website gives ChatGPT, Perplexity, Claude, Gemini and Google's AI Overviews something worth quoting, and fix the pages that don't. Use when the user says or asks \"is my site ready for AI search\", \"audit my site for AI answers\", \"AEO audit\", \"GEO audit\", \"why doesn't ChatGPT quote our pricing\", \"make our pages citable by AI\", \"do I need llms.txt\", \"does schema help with ChatGPT\", \"check our structured data\", \"is our content visible without JavaScript\", \"AI SEO audit\", \"optimize our homepage for AI answers\", \"LLM SEO check\", \"why does AI get our product wrong\", \"fix our site for AI Overviews\". Checks fetchability and indexability, answers in the raw HTML, entity facts, schema (including JS-injected markup), headings, dates and llms.txt, then gives at most 10 prioritised fixes and edits the repo when asked. Works without a Revnu account.","https:\u002F\u002Fgithub.com\u002Fbg-szy\u002FTOP-SKILLS\u002Fblob\u002Fmaster\u002Fskills\u002Fmarketplace\u002Frevnu-ai-readiness-audit\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":56},"01M7DXSZHCVTRS620GC3S97JZV","proxy-troubleshooting","Diagnose residential proxy failures, with NodeMaven's gateway as the reference - 407 with correct credentials, 406, 410, 500, a connection that hangs, ERR_TUNNEL_CONNECTION_FAILED in a browser, Playwright or Puppeteer ignoring proxy credentials, a sticky session that changes IP, geo-targeting that is silently ignored, passwords with special characters, and unexpected traffic from headless Chrome. Use when a proxy connection fails, returns the wrong location, or costs more traffic than expected.","https:\u002F\u002Fgithub.com\u002Fbg-szy\u002FTOP-SKILLS\u002Fblob\u002Fmaster\u002Fskills\u002Fmarketplace\u002Fproxy-troubleshooting\u002FSKILL.md",[],[]]