Aave Liquidation Monitor

Proactive monitoring of Aave V3 borrow positions with liquidation alerts. Queries user collateral, debt, and health factor across chains (Ethereum, Polygon, Arbitrum, etc.). Sends urgent alerts to Telegram/Discord/Slack when health factor drops below configurable thresholds (critical at 1.05, warning at 1.2). Use when you need continuous monitoring of Aave positions, want alerts before liquidation risk occurs, or need periodic summaries of your borrowing health.

Security Scan

The skill largely matches its stated purpose (Aave read-only monitoring) but has a few inconsistencies—most importantly a mismatch between the claimed logging behavior and the code's console output—that the user should review before installing.

Purpose & Capability

Name, description, docs and code align: the code queries Aave public GraphQL endpoints, computes health factor and formats alerts for routing via OpenClaw. No unrelated services, binaries, or credentials are requested in the manifest.

Instruction Scope

SKILL.md and SECURITY.md state 'no sensitive logging' and that OpenClaw handles messaging. However, monitor.js (when run directly) prints the full formatted alert (position details) to console, and cron-runner.js returns the formatted message as well. If OpenClaw captures console output or stores session history, detailed position data could be recorded/logged contrary to the 'no sensitive logging' claim. The instructions also assume OpenClaw will route messages (this is expected) but rely on OpenClaw for secure credential handling; verify OpenClaw's routing and log retention policies.

Install Mechanism

No install spec provided (instruction-only skill with included scripts). No remote downloads or unusual install steps in the repo. The package is distributed as code files—reviewing them locally suffices. This is low install-risk compared to fetching remote executables.

Credentials

No environment variables, API keys, or private keys are requested. The only persistent data is a wallet address stored in OpenClaw prefs (documented). Requested config and runtime access (cron scheduling, OpenClaw messaging) are appropriate for the stated purpose.

Persistence & Privilege

The skill registers cron jobs via OpenClaw's scheduler to run periodically — expected for a monitor. always:false. The scheduler payload uses agentTurn/sessionTarget semantics; nothing in the code modifies other skills or system-wide config. Because cron jobs execute automatically, confirm you are comfortable with background checks and that OpenClaw enforces least privilege for scheduled runs.

What to consider before installing

This skill appears to do what it says (read-only Aave monitoring), but review two things before installing: 1) Logging behavior: SKILL.md/SECURITY.md claim sensitive data and full API responses are not logged. The code prints the full formatted alert message to console in monitor.js (used for manual runs) and cron-runner returns the same message for OpenClaw to route. If OpenClaw captures console output or persists session history, position details may be stored. Confirm OpenClaw's log retention, where console output is recorded, and whether session history is private. 2) Operational assumptions: the skill depends on OpenClaw for secure messaging tokens and cron scheduling. Ensure your OpenClaw instance and messaging integrations (Telegram/Discord/Slack) are configured securely and use a read-only address for monitoring (as recommended). Other notes: the code is otherwise consistent with the docs and uses only Aave's public GraphQL API. If you plan to use it, run a manual /aave-monitor check first, inspect the returned message and OpenClaw logs, and consider auditing the included monitor.js to confirm no additional network calls or logging paths were added in future versions.

Quick Start

Configure your wallet and thresholds once, then the skill runs automatically every 6 hours (configurable).

Step 1: Initialize Your Config

Run /aave-config init to set up:

• Your Ethereum wallet address (read-only; no private keys required)
• Health factor thresholds (default: critical 1.05, warning 1.2, stable >1.5)
• Check interval in hours (default: 6)
• Notification verbosity (verbose = daily summaries even if stable; quiet = only alerts)
• Preferred notification channel (Telegram default, or specify Discord/Slack)

Example:

/aave-config init
→ Wallet: 0x1234...5678
→ Thresholds: 1.05 (critical), 1.2 (warning)
→ Interval: 6 hours
→ Verbosity: quiet (alerts only)
→ Channel: telegram

Step 2: Test the Monitor

Run /aave-monitor check to fetch your position immediately and see the alert format:

/aave-monitor check
→ Fetching position for 0x1234...5678 on Ethereum...
→ Health Factor: 2.31 (Stable ✓)
→ Total Collateral: $50,342.12
→ Total Debt: $21,804.00
→ Borrowed Assets: USDC (15,000), WETH (0.5), USDT (6,804)
→ Supplied Assets: WETH (10), USDC (20,000), DAI (30,000)

Step 3: Enable Proactive Monitoring

Run /aave-monitor enable to start automatic checks every 6 hours. The skill registers a cron job that silently runs in the background and alerts you only when health factor drops below threshold.

For manual override:

/aave-monitor enable --interval 4   # Check every 4 hours
/aave-monitor disable               # Stop monitoring
/aave-monitor status                # See current config + last check result

Configuration Reference

See references/config-guide.md for detailed config options, chain support, and threshold guidance.

See references/aave-api.md for Aave V3 GraphQL schema and query patterns.

Security & Credentials: See SECURITY.md for how credentials are handled and threat model.

For auditors: See SECURITY.md for code review checklist and architecture details.

How It Works

1. Query Phase: Calls Aave V3 GraphQL API with your wallet address
2. Analysis Phase: Extracts health factor, collateral, debt, borrowed/supplied assets
3. Alert Logic:
   • If HF < 1.05 → CRITICAL (red alert, send immediately)
   • If HF < 1.2 → WARNING (yellow alert, send immediately)
   • If HF > 1.5 → STABLE (no alert, unless verbose mode enables daily summary)
4. Notification Phase: Posts alert to your configured channel with clear summary
5. Retry Logic: If API fails, retries up to 3 times with exponential backoff

Error Handling

• Invalid wallet address → Logs error, skips check, retries next interval
• API timeout → Retries up to 3x with exponential backoff (2s, 4s, 8s)
• Malformed response → Alerts you to API changes, logs full response for debugging
• Network errors → Silently retries; alerts only if all retries fail

Security & Credentials

NO private keys are requested or stored — this is read-only monitoring only.

How messaging credentials work

The skill uses OpenClaw's built-in message routing — it does NOT store Telegram, Discord, or Slack tokens.

1. You configure your messaging channel in OpenClaw (outside this skill)
2. The skill calls OpenClaw's messaging API with just the alert text
3. OpenClaw routes the message using your configured channels

Example:

/aave-config set channel telegram
→ OpenClaw sends alerts via your pre-configured Telegram channel

You must have a messaging channel already set up in OpenClaw. See OpenClaw docs for setting up Telegram/Discord/Slack integrations.

Data handling

• Wallet address: Stored in OpenClaw's encrypted config (never sent to external services except Aave's public API)
• API responses: Parsed for health factor only; sensitive position data is NOT logged
• Credentials: Managed by OpenClaw, never stored in skill files

What the skill accesses

✅ Aave GraphQL API (public) — reads your position data
✅ OpenClaw message routing — sends formatted alerts to your configured channel
✅ Cron scheduler — runs background checks on your interval

❌ Does NOT:

• Request private keys
• Store API tokens
• Execute transactions
• Modify your positions

Examples

Alert Examples

Critical Alert (HF < 1.05):

🚨 AAVE LIQUIDATION RISK – CRITICAL
Health Factor: 1.02 (Liquidation threshold: 1.0)
⏰ Action required immediately!

📊 Position Summary:
Total Collateral: $50,000
Total Debt: $48,000
Debt Ratio: 96%

Borrowed (at risk):
- USDC: 40,000
- WETH: 8 ETH (~$31k)

Suggested Actions:
1. Repay debt (especially WETH)
2. Supply more collateral
3. Enable eMode if available (higher LTV on correlated assets)

Last checked: 2026-02-11 08:15 UTC

Warning Alert (HF < 1.2):

⚠️ AAVE WARNING
Health Factor: 1.15 (Approaching 1.2 threshold)

Position approaching liquidation. Consider:
- Repaying $2,000 USDC or
- Adding $5,000 collateral (WETH or stETH)

Stable Position (HF > 1.5, verbose mode only):

✅ AAVE POSITION HEALTHY
Health Factor: 2.31
Total Collateral: $50,342.12
Total Debt: $21,804.00

All clear. Check again tomorrow.

Command Examples

# One-time position check
/aave-monitor check

# Start automatic monitoring
/aave-monitor enable

# Change thresholds (e.g., earlier warning at HF 1.3)
/aave-config set thresholds 1.05 1.3

# Check monitoring status
/aave-monitor status

# View last 5 check results with alerts
/aave-monitor history 5

# Disable monitoring temporarily
/aave-monitor disable

Chains Supported

Primary: Ethereum mainnet

Also works (with same config address where applicable):

• Polygon
• Arbitrum
• Optimism
• Base
• Avalanche
• Gnosis

Edit scripts/monitor.js to add chains or customize per-chain endpoints.

Customization

Change check interval:

/aave-config set interval 4  # Check every 4 hours

Change alert thresholds:

/aave-config set thresholds 1.1 1.25  # Critical at 1.1, warning at 1.25

Enable verbose mode (daily summaries even if stable):

/aave-config set verbosity verbose

Switch notification channel:

/aave-config set channel discord

Implementation Details

See scripts/monitor.js for the core monitoring logic (GraphQL queries, health factor calculation, alert formatting).

See scripts/cron-runner.js for cron job integration with OpenClaw's cron scheduler.

Troubleshooting

"Invalid wallet address" → Check format: must be 0x... format. Try /aave-config view to see current config.

"Health factor not returned by API" → Wallet may not have Aave position. Try /aave-monitor check for immediate response.

"API endpoint unreachable" → Aave API may be down. Skill retries automatically. Check https://status.aave.com.

"No notifications sent" → Check that your notification channel is active. Run /aave-monitor check to test delivery.

Notes

• Health factor = Total Collateral / Total Debt (net of liquidation threshold ratios)
• Liquidation occurs when HF ≤ 1.0
• Variable debt accrues interest; stable debt is fixed-rate
• eMode allows higher LTV for correlated assets (e.g., stablecoins, ETH-LSTs)
• Supplied assets are collateral; they earn interest but can be seized if you're liquidated

---

Last Updated: 2026-02-11 | Status: Ready for configuration