Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Lifi Cross Chain

ASecurity

Use when comparing LI.FI bridge quotes, preparing an unsigned cross-chain transfer, or diagnosing source-confirmed destination-pending, partial or refunded transfers. Reconciles actual destination asset and amount rather than declaring success from a source receipt.

2 stars
0 votes
0 copies
0 views
Added 9/28/2026
ai-agentsnodeapidocumentation

Works with

terminalapi

Security Analysis

A100/100

Scanned 9/28/2026

Install to Claude Code

$npx -y skills add galleonlabs/crypto-defi-skills --skill lifi-cross-chain --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Lifi Cross Chain?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Lifi Cross Chain
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/galleonlabs-lifi-cross-chain/badge)](https://www.skillsdirectory.com/skills/galleonlabs-lifi-cross-chain)

More formats (shields.io, HTML) on the badges page.

Files
SKILL.md
---
name: lifi-cross-chain
description: "Use when comparing LI.FI bridge quotes, preparing an unsigned cross-chain transfer, or diagnosing source-confirmed destination-pending, partial or refunded transfers. Reconciles actual destination asset and amount rather than declaring success from a source receipt."
license: MIT
compatibility: "Read-only EVM RPC and official protocol APIs or UI. Optional Node.js 20 for offline helpers. No signer included."
metadata:
  author: "Galleon Labs"
  version: "0.2.0"
---

# LI.FI Cross-chain

Produce a route comparison with a minimum destination amount and a source-to-destination evidence ledger. For an already submitted route, begin with its source hash and status; do not fetch a new transfer as a recovery action.

## Quote one exact intent

Collect source/destination chain IDs, exact token addresses, raw input amount, sender, recipient, slippage fraction, maximum cost and deadline. Resolve destination gas needs separately from the bridged token. A USDC receipt cannot pay native gas by itself.

Use the official `/v1/quote` recipe in [recipes](references/recipes.md). Quote returns one executable Step with `transactionRequest`; broader `/advanced/routes` comparison yields route steps whose transaction data must be obtained through `/advanced/stepTransaction` when ready. Do not assume a route list already contains executable calldata or precompute later steps from stale balances.

Record quote ID, tool/bridge, included steps, action, `estimate.toAmount`, `toAmountMin`, feeCosts, gasCosts and executionDuration. LI.FI slippage is a fraction: `0.005` means 0.5%. Compare routes for the same assets, recipient, amount and quote window. Compare raw minimum amounts only when destination token and decimals match. Do not subtract fees already reflected in minimum output twice; account separately for extra source/destination native gas.

## Inspect the handoff

Verify `action.fromAddress`, `toAddress`, token/chain identities and returned transaction target/value/data. Check chainId and chain-native gas balance, simulation and quoted allowance. `estimate.approvalAddress` is a spender hint requiring verified protocol/deployment checks, not permission for unlimited approval. `transactionRequest.to` and approval spender need not be the same address.

Use existing official SDK/UI or an authorized wallet tool for requested execution; this skill supplies no signer. Return unsigned terms for research/planning. Exact authority must cover recipient, spend, route, slippage and expiry. If construction or approval changes those terms, re-evaluate before any send. Persist source hash/route ID before waiting for settlement.

## Reconcile until a meaningful terminal outcome

Query `/v1/status` with source hash and chain IDs; bridge/tool can help disambiguate. Observe provider rate limits and retry-after. Do not run an unbounded tight polling loop: report pending with a bounded next-check time when the task cannot wait.

| Provider state | What to report and verify |
| --- | --- |
| NOT_FOUND / INVALID | No indexed transfer yet or bad input; verify source hash/chain/receipt before concluding failure |
| PENDING | Source and destination stages separately; keep same transfer identity |
| DONE + COMPLETED | Verify destination receipt, exact asset, recipient and raw amount against quoted minimum |
| DONE + PARTIAL | Report received asset and amount; destination swap can fail while bridge delivers an intermediate asset |
| DONE + REFUNDED | Identify refund chain, asset, recipient, hash and net amount; original destination obligation is not paid |
| FAILED or unknown | Preserve error/substatus, inspect source/refund evidence; do not create a second transfer automatically |

A provider quote object can appear while pending; it is an estimate, not receiving evidence. A DONE label alone cannot establish intended delivery. Compare `receiving.amount` with the saved `toAmountMin` only after token identity, chain and decimals agree. Reconcile actual destination logs/balances against source debit and both gas costs. A refund is a distinct outcome, not success at the destination.

## Offline outcome classifier

Run `node scripts/settlement.mjs '{"status":"DONE","substatus":"PARTIAL"}'` from this installed skill directory to classify a saved response without network access. For COMPLETED supply positive `expectedChain`, `receivedChain`, EVM `expectedToken`, `receivedToken`, and integer-string `minimumRaw`, `receivedRaw`. The helper compares large amounts exactly and never treats supplied observations as independent chain proof. It intentionally handles EVM token identities only.

## Worked output

Synthetic 1,000 USDC from Base to Arbitrum, minimum 995 USDC. Source receipt succeeds; LI.FI returns DONE/PARTIAL with 0.42 WETH at destination. Report `partial, received 0.42 WETH; 995 USDC objective not satisfied`. Do not compare WETH raw units with a USDC minimum. Show the destination hash and recipient proof; quote a separate WETH-to-USDC remedy only if requested, without repeating the bridge.

For DONE/COMPLETED with actual 997 USDC and quoted estimate 998/minimum 995, report a 1 USDC difference from estimate and 2 USDC above minimum, not slippage failure. Completion remains unverified until the destination evidence agrees.

## Sources and maintenance

Primary sources checked 2026-09-27. [Protocol recipes and sources](references/recipes.md) contain the concrete calls. Recheck deployments and API schemas before preparing financial actions; documentation access alone proves no live position or transaction.

Attribution

galleonlabsgalleonlabs
View sourceMore from galleonlabs →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Ultra-compressed communication mode that cuts output tokens while keeping technical accuracy. Levels: lite, full, ultra and the wenyan variants. Use for /caveman, "caveman mode", "talk like caveman", "be brief" or "less tokens".

1074701 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

695601 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3351 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

691 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →