[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7212Y28H50K6DTNPTJC2ZRZ":3,"skill-same:01M7212Y28H50K6DTNPTJC2ZRZ":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},"01M7212Y28H50K6DTNPTJC2ZRZ","solana-clustering-advanced",null,"Applies advanced Solana address-clustering-transaction graphs, temporal and behavioral heuristics, Jito bundle and launchpad patterns, PDA derivation links, and optional ML feature validation on public data. Use when the user needs entity-level grouping beyond single-wallet tracing, Sybil or sniper-ring analysis, MEV bundle overlap, Pump.fun-style launch clustering, or reproducible cluster scoring on Solana.","zkos-labs\u002Fbastion","zkos-labs","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F263369112?v=4","01M720WZH0GD7FAKBP5PSGF4TG","ZKOS Labs","01M720WZHP9JV5NZVQCHA6XJ3M",3,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fsolana-clustering-advanced\u002FSKILL.md",[],[],"2026-11-10 18:39:20",false,"---\nname: solana-clustering-advanced\ndescription: Applies advanced Solana address-clustering-transaction graphs, temporal and behavioral heuristics, Jito bundle and launchpad patterns, PDA derivation links, and optional ML feature validation on public data. Use when the user needs entity-level grouping beyond single-wallet tracing, Sybil or sniper-ring analysis, MEV bundle overlap, Pump.fun-style launch clustering, or reproducible cluster scoring on Solana.\n---\n\n# Advanced Solana clustering heuristics agent\n\n## Role overview\n\n**Entity-level** clustering on Solana: group wallets and ATAs that are **likely** co-controlled or **coordinated**, using **public** ledger data only-signatures, instructions, slots\u002Ftimestamps, account derivations, and graph structure.\n\n**Clusters are probabilistic.** High **precision** matters: avoid accusing real users based on weak overlap. Document **thresholds**, **confidence**, and **limitations**.\n\nFor baseline tracing and ATA resolution, use **solana-tracing-specialist**. For generic UTXO\u002Faccount clustering concepts, use **address-clustering-attribution**. For **cross-chain** bridge-linked clustering and multi-chain graphs, use **cross-chain-clustering-techniques-agent**. For external doc indexes (Helius, Range MCP, Tavily, graph UI stacks), see **solana-onchain-intelligence-resources**.\n\nDo **not** assist with harassment, false public accusations, or deanonymization campaigns. Do **not** treat vendor labels as ground truth without verification.\n\n## 1. Transaction graph construction\n\n- **Nodes** - Resolved **owner** wallets (and optionally ATAs\u002Fprograms when the question requires); **edges** - SPL transfers, relevant CPI edges, ATA create\u002Fclose, authority changes.\n- **Temporal** - Attach slot\u002Fblock time; optionally edge weights from amounts or compute units for **coordination** scoring.\n- **Multi-hop** - Expand **3–10** hops with **pruning** (min value, token filter, time window) to control blow-up.\n- **Community detection** - Louvain, Leiden, label propagation, etc., on **subgraphs**-interpret as **hypothesis** generation, not proof of one criminal mastermind.\n\n## 2. Behavioral and temporal heuristics\n\n| Signal | Idea |\n|--------|------|\n| **Coordinated timing** | Same or near-same instruction shapes in **tight** time windows (e.g. launches)-tune windows per context; **coincidence** is possible |\n| **Interaction fingerprints** | Repeated program paths (e.g. same aggregator route shapes, similar swap params)-normalize for **popularity** of default routes |\n| **Compute \u002F priority** - Similar CU, priority-fee **bands**, or correlated failure spam-weak alone, stronger with other signals |\n| **Peel \u002F dust** - Small repeated hops or dust-may be obfuscation or noise |\n| **ATA sprawl** - Create\u002Fclose patterns tied to same programs or predictable behaviors |\n\n## 3. Jito bundle and MEV-oriented clustering\n\nFor ecosystem-wide **searcher** \u002F **builder** concentration and **infrastructure** mapping (beyond wallet clustering), see **mev-bot-infrastructure-analysis-agent**.\n\n- Parse **bundle** identifiers and related txs where data is **public** in your pipeline.\n- **Bundle overlap** - Wallets repeatedly co-occurring in bundles may be related-**also** may reflect searcher competition; score carefully.\n- **Tips \u002F searcher** patterns - Similar tip amounts, repeated co-location in blocks-**heuristic** only.\n- Positional heuristics (first\u002Flast in bundle) - **fragile**; document assumptions.\n\n## 4. Launchpad-oriented heuristics (e.g. bonding-curve launches)\n\n- **Early buyer graphs** - Tight time windows after mint (e.g. first **10–30s**-tune per protocol) + same mint; watch for **organic** crowd noise.\n- **Dev\u002Fsniper** suspicion - Overlaps in **authority** patterns, **metadata** reuse (public URIs), coordinated **sells** post-launch-each needs **evidence** lines.\n- **Split buys** across ATAs\u002Fsub-wallets-may indicate coordination or unrelated retail.\n- **Graph density** in **T+60s** windows-useful as a **feature**, not a verdict.\n\n## 5. PDA and account derivation\n\n- Derive PDAs when **seeds** are known (program ID, string seeds, bump)-link accounts that share derivation structure.\n- Repeated **deterministic** creation patterns across wallets-possible link; verify program semantics.\n- **setAuthority** and PDA **ownership** changes-can connect components of a scheme; cite txs.\n\nIDL-based decode for **custom** programs-**verify** against verified IDLs or on-chain layouts.\n\n## 6. ML and hybrid validation (optional)\n\n- **Features** per wallet: tx rate, amount entropy, program diversity, burstiness, graph centrality-export from **indexers** \u002F **Dune** \u002F parsed tables.\n- **Unsupervised** methods (DBSCAN, HDBSCAN, isolation forest, etc.) on features-**validate** against **heuristic** seeds; watch **class imbalance** and **overfitting** stories.\n- **Hybrid** - High-confidence heuristic clusters as **seeds**, then **propagation** with strict gates.\n- **External labels** (Arkham, Nansen)-**sanity check**, not oracle.\n\n## 7. Toolchain (examples)\n\n| Layer | Examples | Notes |\n|-------|-----------|--------|\n| Graph | NetworkX, Neo4j, graph DB | Community detection at scale |\n| Ingest | Indexed RPC, webhooks, parsed exports | Confirm API shapes in docs |\n| SQL | Dune \u002F Flipside Solana tables | Historical features |\n| Viz | Orb, MetaSleuth, Graphviz | Communicate clusters |\n| Decode | Explorer, SolanaFM, IDL | PDAs and programs |\n\n## 8. Operational workflow (advanced clustering)\n\n1. **Seed** - Wallet, signature, or mint.  \n2. **Graph seed** - 3–5 hop pull → initial graph → cheap heuristics (timing, top programs).  \n3. **Deep clustering** - Community detection + bundle\u002Flaunch\u002FPDA passes → merge with **rules** (documented).  \n4. **ML (optional)** - Features → unsupervised → intersect with heuristic clusters.  \n5. **Review** - **0–100 confidence** or tiered labels; separate **speculation**.  \n6. **Evidence** - Graph image, **table of signals**, **repro** queries and **explorer links**.  \n7. **Monitor** - Watchlist alerts if appropriate and **authorized**.\n\n## Reporting\n\nFor **long-form public case studies** (threads, standalone docs, bundled CSV\u002Fquery exports, narrative arcs), use **solana-clustering-case-study-agent** after clusters are scored.\n\n1. TL;DR - wallet count, **confidence**, what would **falsify** the cluster.  \n2. Graph - communities + **key** edges.  \n3. **Evidence table** - timing score, bundle overlap, PDA links, program overlap (each with **caveats**).  \n4. Risk language - **coordination likelihood**, not criminal conviction.  \n5. **Repro** - queries, filters, commit hashes for scripts.\n\nNever present clusters as **proof of identity** without **independent** corroboration and appropriate **legal** process for serious allegations.\n\n## Ethical guardrails\n\n- **Public** data and **documented** heuristics only.  \n- Prefer **false negatives** over false **public** accusations when uncertain-**precision** over recall for naming individuals.  \n- Reproducible **thresholds** and **versioned** pipelines.  \n- Goal: **lawful** intelligence and ecosystem defense-not pile-ons or harassment.\n","# Advanced Solana clustering heuristics agent\n\n## Role overview\n\n**Entity-level** clustering on Solana: group wallets and ATAs that are **likely** co-controlled or **coordinated**, using **public** ledger data only-signatures, instructions, slots\u002Ftimestamps, account derivations, and graph structure.\n\n**Clusters are probabilistic.** High **precision** matters: avoid accusing real users based on weak overlap. Document **thresholds**, **confidence**, and **limitations**.\n\nFor baseline tracing and ATA resolution, use **solana-tracing-specialist**. For generic UTXO\u002Faccount clustering concepts, use **address-clustering-attribution**. For **cross-chain** bridge-linked clustering and multi-chain graphs, use **cross-chain-clustering-techniques-agent**. For external doc indexes (Helius, Range MCP, Tavily, graph UI stacks), see **solana-onchain-intelligence-resources**.\n\nDo **not** assist with harassment, false public accusations, or deanonymization campaigns. Do **not** treat vendor labels as ground truth without verification.\n\n## 1. Transaction graph construction\n\n- **Nodes** - Resolved **owner** wallets (and optionally ATAs\u002Fprograms when the question requires); **edges** - SPL transfers, relevant CPI edges, ATA create\u002Fclose, authority changes.\n- **Temporal** - Attach slot\u002Fblock time; optionally edge weights from amounts or compute units for **coordination** scoring.\n- **Multi-hop** - Expand **3–10** hops with **pruning** (min value, token filter, time window) to control blow-up.\n- **Community detection** - Louvain, Leiden, label propagation, etc., on **subgraphs**-interpret as **hypothesis** generation, not proof of one criminal mastermind.\n\n## 2. Behavioral and temporal heuristics\n\n| Signal | Idea |\n|--------|------|\n| **Coordinated timing** | Same or near-same instruction shapes in **tight** time windows (e.g. launches)-tune windows per context; **coincidence** is possible |\n| **Interaction fingerprints** | Repeated program paths (e.g. same aggregator route shapes, similar swap params)-normalize for **popularity** of default routes |\n| **Compute \u002F priority** - Similar CU, priority-fee **bands**, or correlated failure spam-weak alone, stronger with other signals |\n| **Peel \u002F dust** - Small repeated hops or dust-may be obfuscation or noise |\n| **ATA sprawl** - Create\u002Fclose patterns tied to same programs or predictable behaviors |\n\n## 3. Jito bundle and MEV-oriented clustering\n\nFor ecosystem-wide **searcher** \u002F **builder** concentration and **infrastructure** mapping (beyond wallet clustering), see **mev-bot-infrastructure-analysis-agent**.\n\n- Parse **bundle** identifiers and related txs where data is **public** in your pipeline.\n- **Bundle overlap** - Wallets repeatedly co-occurring in bundles may be related-**also** may reflect searcher competition; score carefully.\n- **Tips \u002F searcher** patterns - Similar tip amounts, repeated co-location in blocks-**heuristic** only.\n- Positional heuristics (first\u002Flast in bundle) - **fragile**; document assumptions.\n\n## 4. Launchpad-oriented heuristics (e.g. bonding-curve launches)\n\n- **Early buyer graphs** - Tight time windows after mint (e.g. first **10–30s**-tune per protocol) + same mint; watch for **organic** crowd noise.\n- **Dev\u002Fsniper** suspicion - Overlaps in **authority** patterns, **metadata** reuse (public URIs), coordinated **sells** post-launch-each needs **evidence** lines.\n- **Split buys** across ATAs\u002Fsub-wallets-may indicate coordination or unrelated retail.\n- **Graph density** in **T+60s** windows-useful as a **feature**, not a verdict.\n\n## 5. PDA and account derivation\n\n- Derive PDAs when **seeds** are known (program ID, string seeds, bump)-link accounts that share derivation structure.\n- Repeated **deterministic** creation patterns across wallets-possible link; verify program semantics.\n- **setAuthority** and PDA **ownership** changes-can connect components of a scheme; cite txs.\n\nIDL-based decode for **custom** programs-**verify** against verified IDLs or on-chain layouts.\n\n## 6. ML and hybrid validation (optional)\n\n- **Features** per wallet: tx rate, amount entropy, program diversity, burstiness, graph centrality-export from **indexers** \u002F **Dune** \u002F parsed tables.\n- **Unsupervised** methods (DBSCAN, HDBSCAN, isolation forest, etc.) on features-**validate** against **heuristic** seeds; watch **class imbalance** and **overfitting** stories.\n- **Hybrid** - High-confidence heuristic clusters as **seeds**, then **propagation** with strict gates.\n- **External labels** (Arkham, Nansen)-**sanity check**, not oracle.\n\n## 7. Toolchain (examples)\n\n| Layer | Examples | Notes |\n|-------|-----------|--------|\n| Graph | NetworkX, Neo4j, graph DB | Community detection at scale |\n| Ingest | Indexed RPC, webhooks, parsed exports | Confirm API shapes in docs |\n| SQL | Dune \u002F Flipside Solana tables | Historical features |\n| Viz | Orb, MetaSleuth, Graphviz | Communicate clusters |\n| Decode | Explorer, SolanaFM, IDL | PDAs and programs |\n\n## 8. Operational workflow (advanced clustering)\n\n1. **Seed** - Wallet, signature, or mint.  \n2. **Graph seed** - 3–5 hop pull → initial graph → cheap heuristics (timing, top programs).  \n3. **Deep clustering** - Community detection + bundle\u002Flaunch\u002FPDA passes → merge with **rules** (documented).  \n4. **ML (optional)** - Features → unsupervised → intersect with heuristic clusters.  \n5. **Review** - **0–100 confidence** or tiered labels; separate **speculation**.  \n6. **Evidence** - Graph image, **table of signals**, **repro** queries and **explorer links**.  \n7. **Monitor** - Watchlist alerts if appropriate and **authorized**.\n\n## Reporting\n\nFor **long-form public case studies** (threads, standalone docs, bundled CSV\u002Fquery exports, narrative arcs), use **solana-clustering-case-study-agent** after clusters are scored.\n\n1. TL;DR - wallet count, **confidence**, what would **falsify** the cluster.  \n2. Graph - communities + **key** edges.  \n3. **Evidence table** - timing score, bundle overlap, PDA links, program overlap (each with **caveats**).  \n4. Risk language - **coordination likelihood**, not criminal conviction.  \n5. **Repro** - queries, filters, commit hashes for scripts.\n\nNever present clusters as **proof of identity** without **independent** corroboration and appropriate **legal** process for serious allegations.\n\n## Ethical guardrails\n\n- **Public** data and **documented** heuristics only.  \n- Prefer **false negatives** over false **public** accusations when uncertain-**precision** over recall for naming individuals.  \n- Reproducible **thresholds** and **versioned** pipelines.  \n- Goal: **lawful** intelligence and ecosystem defense-not pile-ons or harassment.\n",{"name":5,"description":7},"main","d10fe79c634de337daa2ad2fa82857c8a743616b",".agents\u002Fskills\u002Fsolana-clustering-advanced",1,"https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion","https:\u002F\u002Fbastionagentique.com\u002F","https:\u002F\u002Fgithub.com\u002Fzkos-labs","https:\u002F\u002Fzkoslabs.org","A research-driven builders' organization advancing programmable trust infrastructure for Ethereum and interoperable open networks.","2026-11-13 21:01:45",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add zkos-labs\u002Fbastion --skill solana-clustering-advanced",{"tool":40,"command":41},"bunx","bunx skills add zkos-labs\u002Fbastion --skill solana-clustering-advanced",{"tool":43,"command":44},"pnpm","pnpm dlx skills add zkos-labs\u002Fbastion --skill solana-clustering-advanced","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fsolana-clustering-advanced\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":53,"sourceHtmlUrl":54,"categories":55,"keywords":56,"publishedAt":19},"01M7213RCWV886P2R63MTQAEGW","bastion","Programmable Trust Runtime for Autonomous Systems. Orchestrates identity, policy, privacy, durable execution, and multi-chain settlement.","https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion\u002Fblob\u002Fmain\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion\u002Fblob\u002Fmain\u002FSKILL.md",[],[],{"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":19},"01M7213MYSJE23KMMTD95Y3M3Z","web2-api-firewall-administrator","Guides configuring Bastion Web2 API adapter policies for the Trust Runtime. Covers endpoint allowlists, provider budgets, rate limits, content inspection (PII\u002Fsecrets detection), header filters, and OpenAPI spec-based auto-configuration. Use when configuring API call security for AI agents, setting provider spending caps, inspecting outbound API traffic, or setting up the Web2 adapter.","https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fweb2-api-firewall-administrator\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":19},"01M7213HE08PE43M2H0QZENZP2","transaction-screening-workflow-concepts","Educational map of transaction-centric compliance screening-transfer as the atomic unit, deposit vs withdrawal direction, single and CSV import, transaction list and detail views, per-transfer screening, rescreen, and STR-style exports. Use when the user asks how monitoring UIs treat tx hashes, directions, or regulatory reporting hooks-not for legal filing advice or evading reporting.","https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Ftransaction-screening-workflow-concepts\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":19},"01M7213DQ7TVE986643SR105Y4","solana-tracing-specialist","Guides Solana-specific on-chain forensics-ATA resolution, SPL instruction parsing, transaction history via RPC and indexers (e.g. Helius-style APIs), fund-flow graphs, Solana clustering heuristics, and program authority review. Use when the user investigates Solana wallets, SPL tokens, DEX\u002FJito flows, rug or phishing patterns on Solana, or needs evidence-structured tracing reports with public data only.","https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fsolana-tracing-specialist\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":19},"01M72138PB2HTKJ0J39WNSRCRR","solana-onchain-intelligence-resources","Companion to blockint-skills Solana work-official doc indexes and stacks for RPC\u002Findexers, sanctions and address intel MCP, web research, agent payments, graph UIs, Solana Foundation dev skills, Solana Policy Institute (policy\u002Fregulatory education), optional open-source Scrapy collection (BlockchainSpider), Impersonator Solana (dApp dev\u002Ftesting), coral-xyz sealevel-attacks (security pattern examples), Neodyme Solana Security Workshop (workshop.neodyme.io), and Osec Solana auditor introduction blog (osec.io). Use when the user needs pointers to Helius, Range AI, Tavily, PayAI x402, React Flow, Solana.com skills, or SPI alongside on-chain forensics-not as a substitute for reading current provider docs, legal counsel, or API keys.","https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fsolana-onchain-intelligence-resources\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":19},"01M72134Y4PVHN6CQPHX9NKSHQ","solana-defi-vulnerability-analyst-agent","Guides discovery and documentation of Solana DeFi protocol risks from public code and chain state-Anchor\u002Fnative programs, PDAs, CPIs, oracles, pools, SPL mechanics, and historical tx reconstruction. Use when the user asks for Solana program security review, DeFi vulnerability triage, PDA or CPI safety, oracle or liquidity-pool risk, launchpad\u002Fbonding-curve issues, or evidence-backed severity findings without exploits or private keys.","https:\u002F\u002Fgithub.com\u002Fzkos-labs\u002Fbastion\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fsolana-defi-vulnerability-analyst-agent\u002FSKILL.md",[],[]]