Borrowing it
Nothing to install: this file belongs to 0xfreak0/sui-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/0xfreak0/sui-mcp/main/.claude/skills/sui-forensics/SKILL.mdgit clone --depth 1 https://github.com/0xfreak0/sui-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/0xfreak0/sui-mcp/sui-forensics)<a href="https://agentmods.dev/skills/0xfreak0/sui-mcp/sui-forensics"><img src="https://agentmods.dev/badge/skills/0xfreak0/sui-mcp/sui-forensics/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/0xfreak0/sui-mcp/sui-forensics"><img src="https://agentmods.dev/badge/skills/0xfreak0/sui-mcp/sui-forensics.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00074 | $0.02755 |
| Opus 5 | $0.00037 | $0.01378 |
| Sonnet 5 | $0.00015 | $0.00551 |
| Haiku 4.5 | $0.00007 | $0.00276 |
Grade A, and why
sui-forensics scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 217 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Investigating on Sui
The hard part of this work is not fetching data. It is knowing what the data does not say. Chain data is complete and public, which makes a wrong conclusion look exactly like a right one — fluent, specific, and sourced.
Everything below exists because the plain reading was wrong at least once.
Evidence tiers, and what each licenses
Every claim should be traceable to one of these. Say which.
| tier | means | you may write |
|---|---|---|
chain-derived |
Read from Sui itself | "X sent 5 SUI to Y in transaction Z" |
indexer-attested |
A third party asserts it | "Wormholescan reports this VAA was redeemed on Ethereum" — a lead to confirm, not a finding |
heuristic |
An inference from patterns | "These addresses may share an operator" — never "they do" |
build_wallet_edges is the only tool that tags its output heuristic. Its
edges are facts with digests attached; its clusters are inference. Do not
collapse the two, and never record a cluster as a finding without confirming it
yourself.
Opening a case
- Set your sinks first. Labels decide where a trace stops. Curated ones
ship nearly empty on purpose — attribution is case-specific and mostly not
publishable — so add what you know with
manage_labels, or pointSUI_LABELS_FILEat a private file for a whole case. A trace that runs past a known exchange, or stops at one you never told it about, is usually this. - Identify before you trace.
identify_address. A hop that is a package or a shared object is not "someone the funds went to", and a trace that treats a DEX pool as a person is wrong from that point on. - Trace with
trace_funds. Readstop_reasonandunfollowedbefore the path: a trace follows one branch, and splitting across wallets is the ordinary laundering move. - Attribute with
find_funding_source— orfind_funding_sourcesfor several addresses at once, which also reports co-funding and its denominators. Then measure the funder withget_address_fanoutbefore believing anything. - Cluster only once you have a reason to.
build_wallet_edgesanswers "is this a new party or the same one", not "who is this". - Record with
save_finding, one claim per finding, with its digests.export_casewhen the case outlives the session.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago Changed · +95 lines 31513ebc1f4c
- 5d ago First seen · 122 lines · 74 tokens per session scan A ade0332cd795
sui-forensics is a skill published in the GitHub repository 0xfreak0/sui-mcp (1 stars, last pushed today), licensed MIT. It adds 74 tokens to every session and 2,755 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-05.
Other skills, from other repositories
openclaw-crypto-payments
Skill "openclaw-crypto-payments" from yedanyagamiai-cmd/openclaw-mcp-servers, covering openclaw crypto payments v2.0, what's new in v2.0, quick start, detection triggers and named protocols.
algorand-interaction
Interact with Algorand blockchain via the Algorand MCP server — wallet operations, ALGO/ASA transactions, smart contracts, account info, NFD lookups, atomic groups, Tinyman swaps, Haystack Router best-price swaps, Alpha Arcade prediction markets, Pera asset verification, TEAL compilation, knowledge base. Use when user…
web3-developer
Activates Web3-Developer for smart contract development and DeFi protocol engineering. Use when you need gas-optimized Solidity smart contracts, AMM or lending protocol architecture, reentrancy/overflow/flash loan attack detection in smart contracts, Hardhat/Foundry test suite design with fork testing, or…
nft-standards
Implement NFT standards (ERC-721, ERC-1155) with proper metadata handling, minting strategies, and marketplace integration. Use when creating NFT contracts, building NFT marketplaces, or implementing digital asset systems.
web3-testing
Test smart contracts comprehensively using Hardhat and Foundry with unit tests, integration tests, and mainnet forking. Use when testing Solidity contracts, setting up blockchain test suites, or validating DeFi protocols.
delegacion-local
Regla y catálogo para delegar pasos mecánicos (resumir, clasificar, extraer, boilerplate, mensaje de commit desde un diff, traducir texto o archivo, resumir salida de lint/tests/CI, explicar código, describir una imagen, verificar si el backend local está disponible) a modelos locales vía las tools local del MCP…