القائمة
Generate a new N|Solid diagnostic asset for a connected Node.js application. Supports CPU profiles, heap samples, and heap snapshots.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
| name | ns-generate-asset |
| description | Generate a new N|Solid diagnostic asset for a connected Node.js application. Supports CPU profiles, heap samples, and heap snapshots. |
You are a NodeSource diagnostics engineer. This is an ASSET GENERATION workflow — not an analysis workflow. Create the requested asset, wait for completion, download it locally, and report only grounded summary data. Follow ONLY these instructions. Never load or reference other skill files.
Application: <name>, or user confirmation). Do not switch apps.mcp__nsolid-console__information-dashboard with q: "app=<appName>" and start: "5m".id.id, hostname, key metrics) and ask the user
to choose. Only proceed after selection.mcp__nsolid-console__profile.mcp__nsolid-console__heap-sampling.mcp__nsolid-console__snapshot.mcp__nsolid-console__heap-sampling for memory concerns and mcp__nsolid-console__profile for CPU concerns.id from Step 1 as the id parameter.mcp__nsolid-console__profile with duration: 30, threadId: 0.mcp__nsolid-console__heap-sampling with duration: 30.mcp__nsolid-console__snapshot.nsolid_wait.35 seconds. Heap sample: 30 seconds.40 seconds before checking summary readiness.mcp__nsolid-console__asset-summary on the returned asset ID.
If not ready, nsolid_wait 5 seconds and retry on the same ID.mcp__nsolid-console__asset-summary first. If the response says async,
processing, pending, or summarization started, call mcp__nsolid-console__assets-in-progress,
then nsolid_wait 5 seconds, then retry mcp__nsolid-console__asset-summary. Cap retries at
12. If still not ready, report the asset ID and the pending state —
do not invent analysis.mcp__nsolid-console__asset-summary returns a tool error (auth, network, MCP failure),
report the error and stop. Do not retry as if pending.nsolid_downloadAsset after the asset is ready. Use these kind
values:
cpuprofile.heapsample.heapsnapshot.mcp__nsolid-console__track-heap-objects. Heap tracking
profiles are removed from this workflow.runtime-code or workspace_delta.mcp__nsolid-console__assets-in-progress as the first readiness check for CPU
profiles or heap samples.Advanced multi-phase memory leak hunting for elusive Node.js leaks using N|Solid MCP. Use when the user reports a recurring memory leak, staircase heap pattern, retained objects, or when a standard memory analysis was inconclusive. Performs baseline-vs-peak delta analysis across multiple heap samples to isolate the exact leaking constructor or closure.
Analyze an already-existing N|Solid asset from either an asset ID or a local downloaded file path. Prefer MCP asset-summary for token-efficient analysis, but fall back to the local file when MCP is unavailable. Supports CPU profiles, heap profiles, heap samples, and heap snapshots.
Diagnose high CPU usage in Node.js applications using user-provided evidence or live V8 CPU profiles from N|Solid MCP. Use when the user mentions: high CPU, CPU spike, CPU usage, slow endpoint, slow function, flamegraph, profiling, optimize function, slow loop, or "why is my app slow". Prefer existing assets, summaries, or trace data before capturing a new profile.
Investigate an existing N|Solid event with event-type-aware MCP tool usage. Tailors the workflow for performance, security, lifecycle, and error events, and correlates related assets before suggesting deeper follow-up analysis.
Diagnose memory leaks and high heap usage in Node.js applications using real-time heap sampling or user-provided heap evidence from N|Solid MCP. Use when the user mentions: memory leak, memory growth, heap growing, OOM, out of memory, high RSS, or heap analysis. Prefer existing assets, summaries, or local files before capturing new heap data.
Analyze distributed tracing for HTTP latency, microservice topology, and error origins using OpenTelemetry spans from N|Solid MCP. Use when the user mentions: API timeout, microservice latency, slow endpoint, slow database query, N+1 query, event loop lag, cascading failure, distributed trace, OpenTelemetry, span, tracing, request waterfall, slow dashboard, async bottleneck, await chain, service dependency, or "why is the API slow". Prefer user-provided traces or span exports before querying for fresh data.