Works with the AI Agents you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+21 moreClaude CodeCursorCodex CLIMuseDots
Manus
Grok
Grok Bot+18 agentsBlockchain RPC Reliability Diagnostic Agent
Diagnose blockchain RPC timeouts, rate limits, and latency issues with structured log analysis and remediation plans.
$39.99
Blockchain RPC Reliability Diagnostic Agent
Example session with this skill installed
I am seeing intermittent 429s and high p99 latency on eth_getLogs calls using Alchemy. Here are my sanitized logs from the last 10 minutes and my current Ethers.js retry config.
- Read your context and instructions
- Compiled the blockchain rpc reliability
I have analyzed the provided logs and configuration for your Ethereum mainnet RPC integration. The evidence suggests a combination of aggressive burst patterns and unoptimized query ranges.
RPC Reliability Diagnostic Report
1. Incident Summary
- Chain: Ethereum Mainnet
- Method:
eth_getLogs - Symptom: Intermittent HTTP 429 (Too Many Requests) and p99 latency spikes (>2500ms).
- Timeframe: 10-minute window.
2. Error and Rate-Limit Analysis
| Metric/Error | Observation | Confidence |
|---|---|---|
| HTTP 429 | 14 occurrences detected during 2-second bursts. | High |
| JSON-RPC Error | None. Errors are transport-level rate limits. | High |
| Burst Pattern | Client is sending 50+ concurrent log requests every 30s. | Medium |
3. Latency Findings
Logs show a strong correlation between block range size and latency. Queries spanning >100 blocks exhibit a p99 of 2.8s, while 10-block ranges stay under 400ms.
4. Retry-Policy Review
Current config uses linear backoff (1000ms). This is contributing to "retry storms" where failed requests all retry simultaneously, hitting the rate limit again immediately.
5. Prioritized Remediation Plan
Immediate: Switch to Exponential Backoff with Jitter. Implementation of fixed 1s to 2^n + random_jitter will alleviate burst collisions.
2. Optimization: Reduce eth_getLogs block range from 100 to 20 blocks per call.
3. Architecture: Implement a request queue to cap concurrency at 10 active requests.
Next steps
- Update your Ethers.js
StaticJsonRpcProviderwith a customExponentialBackoffcollector. - Benchmark the 20-block range latency compared to the current 100-block range.
- Verify if Alchemy "Compute Units" (CUs) are being exhausted by the specific filter topics used.
Connects securely to your tools. The creator never sees your data.
About this skill
Debugging intermittent blockchain RPC failures usually involves chasing ghost logs and guessing at provider-side rate limits. This skill provides a structured diagnostic framework to analyze RPC logs, metrics, and configurations for EVM and non-EVM chains. It isolates transport layers from application-level JSON-RPC errors to stop the trial-and-error cycle of switching providers.
What it does
- Error Categorization distinguishes between DNS/TLS transport failures, HTTP status codes, and JSON-RPC application errors like execution reverts.
- Latency Bottleneck Analysis reviews p95/p99 metrics and method mixes to identify query complexity or concurrency issues.
- Rate-Limit Auditing detects burst patterns and 429 responses to estimate necessary request pacing and backoff jitter.
- Retry Policy Review evaluates maximum attempts and idempotency-aware handling for transaction submissions versus read-only calls.
- Failover Architecture designs circuit breaker logic and health check strategies to prevent provider-switching loops.
How it works
- Submit Evidence by providing redacted logs, client library versions, and current retry configurations.
- Analysis Phase where the agent groups logs by method, status code, and provider to find correlations.
- Hypothesis Generation resulting in a diagnostic report labeled by confidence levels (High, Medium, Low).
- Remediation Implementation following a prioritized plan with specific test cases for each suggested fix.
Frameworks & tools
Works with any blockchain client library including Ethers.js, Viem, Web3.js, and Rust-based clients. Compatible with logs from providers like Alchemy, Infura, QuickNode, or self-hosted nodes.
Why this beats prompting it yourself
Generic prompts often fail to distinguish between transient network errors and permanent JSON-RPC failures. This skill enforces a strict diagnostic workflow that separates observations from hypotheses, preventing "hallucinated" latency bottlenecks and ensuring retry logic doesn't cause transaction duplication.
Use cases
- Debugging frequent 429 Too Many Requests errors during high-volume indexing.
- Optimizing p99 latency for real-time MEV or trading bot execution.
- Designing a multi-provider failover strategy for production dApps.
- Auditing transaction submission logic to prevent "nonce already used" errors during spikes.
Known limitations
Cannot perform live network probes or active endpoint testing. Requires users to manually supply logs and metrics. Accuracy depends on the quality and timeframe of provided log data.
Cannot perform live network probes. Requires manually supplied logs. Accuracy depends on provided data quality. Does not handle private key or credential management.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 22 days ago