# Crypto Intelligence/onchain > House of On-Chain skill. Auto-activates on /crypto-onchain-* commands or work within verticals/crypto-intelligence/onchain/. Enforces R5 non-advisory clause inline, chain-data source citation, mechanism-context grounding, named output artifact discipline. - Skill: `frankxai/crypto-intelligence-onchain` (Agent Skill, multi-file: 9 files) - Install (CLI): `npx skillmds@latest add frankxai/crypto-intelligence-onchain` - Raw SKILL.md: https://api.skillmd.com/api/skills/frankxai/crypto-intelligence-onchain/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Research & Search - Author: frankxai (https://skillmd.com/u/frankxai) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/frankxai/crypto-intelligence-onchain --- # SKILL — Crypto Intelligence / House of On-Chain (House-tier) > House-tier skill. Inherits Crypto Intelligence vertical-tier skill rules (R5 clause inline, voice = architect, refusal patterns, composition with Wealth IS umbrella). Adds House-of-On-Chain-specific discipline. --- ## R5 non-advisory clause (mandatory inline, every artifact) ``` > This is system architecture, not financial / investment / tax / legal advice. > Outputs frame decisions; jurisdiction-specific counsel signs off on instruments. > The practitioner accepts capital risk; the substrate accepts no claim. > Cryptocurrency markets carry total-loss risk; this vertical does not represent > recovery probability or expected return. ``` This block opens every House of On-Chain output artifact. Non-waivable. Audited per Crypto IS SOUL.md drift test 1. --- ## House-specific reasoning protocol ### Step 1 — Declare scope Name explicitly: - Which wallets / protocols / networks / tx-hashes are in-scope - What time window (rolling 7d / 30d / cycle-window / since-event) - What sister-House this artifact feeds (Allocation sizing / Macro regime / Sovereignty custody-review / Research thesis) ### Step 2 — Cite chain-data sources Every claim traces to a named source: - Dune Analytics: query ID + dashboard URL - Etherscan / Solscan / equivalent: tx-hash + block height - Nansen: wallet-label + entity tag + last-update timestamp - Arkham: entity attribution + confidence band - DefiLlama: protocol slug + snapshot timestamp - Flashbots / MEV-Inspect / Eigenphi: query method + window No "the data shows" without source. No "wallet activity suggests" without entity-tag source. No "TVL is" without DefiLlama-or-equivalent snapshot timestamp. ### Step 3 — Cite mechanism context For each on-chain action analyzed, name the mechanism: - Swap routing (Uniswap V3 vs CoW vs 1inch vs MEV-Protect — different mechanisms, different MEV-exposure profiles) - Bridge type (lock-and-mint vs burn-and-mint vs liquidity-network — different risk-stacks) - Validator-econ model (per-network, no generalized "PoS" framing) - Governance vote (token-weighted vs delegated vs quadratic — different attack surfaces) ### Step 4 — Flag interpretive uncertainty Wallet-attribution is often ambiguous. MEV-vs-arbitrage classification is contested. Validator slashing data has reporting lags. Flag where the analysis is uncertain — do not paper over with concrete-sounding numbers. ### Step 5 — Ship the named output artifact Every command ships a named file: - `/crypto-onchain-flow-snapshot` → `flow-snapshot-.md` (or instance-private equivalent) - `/crypto-onchain-wallet-trace` → `wallet-trace--.md` - `/crypto-onchain-mev-audit` → `mev-audit--.md` - `/crypto-onchain-validator-econ` → `validator-econ--.md` - `/crypto-onchain-contract-interaction` → `contract-trace-.md` The filename is the contract. If the file isn't written, the command didn't run. --- ## Refusal patterns (House-specific, inherited from Crypto IS SOUL.md) This House refuses: 1. **"Wallet X is buying — bullish" theses without protocol-mechanism grounding.** Wallet activity is one signal; conclusions require mechanism analysis from House of Research. 2. **MEV-extraction-from-retail plays.** Defensive MEV-audit only. The House produces *exposure* characterizations, not *extraction* strategies. 3. **Sybil-airdrop-farming optimization.** Algorithmic-gaming corruption mode. 4. **Wash-trading volume as legitimacy.** Refused. 5. **Concrete-sounding-stat-without-source.** Universal corruption mode (per Crypto IS SOUL + Wealth IS + Sound IS + _template). 6. **Whale-attribution without entity-tag source.** Anonymous wallet → unnamed actor. No "this wallet belongs to X" without Nansen/Arkham tag citation. --- ## Composition with sister Houses + Wealth IS umbrella | Composes with | Direction | Cite as | |---|---|---| | House of Macro (v0.2) | Flow snapshots feed regime-call inputs | `Feeds: /crypto-macro-regime` | | House of DeFi (v0.2) | Protocol-interaction data feeds mechanism audit | `Feeds: /crypto-defi-mechanism-audit` | | House of Sovereignty (v0.2) | Activity patterns feed custody-tier review | `Feeds: /crypto-sov-custody-design` | | House of Research (v0.2) | On-chain activity surfaces tokenomics signals | `Feeds: /crypto-res-tokenomics` | | House of Allocation (v0.2) | Whale-tracking + flow → sizing inputs | `Feeds: /crypto-alloc-sizing` | | Wealth IS umbrella | Concentration risk inputs to cross-asset review | `Feeds: /wealth-portfolio-fit` | Each artifact declares its composes-with downstream explicitly. --- ## Quality gates (must pass before artifact ships) - [ ] R5 non-advisory clause inline at top of artifact - [ ] Chain-data source(s) cited (Dune ID / tx-hash / Nansen label / Arkham entity / DefiLlama slug) - [ ] Mechanism context provided for each on-chain action - [ ] Interpretive uncertainty flagged where data is ambiguous - [ ] Composes-with declared if artifact feeds sister House - [ ] Named output artifact written to canonical path - [ ] "Built on SIP" attestation block at footer (ambient per v7.4) If any gate fails → artifact does not ship → re-execute with gate satisfied. --- **Built on SIP** — Crypto / House of On-Chain skill (House-tier) · v0.1 proof-of-pattern · SIP v1.1.0