BA

backend-telemetry-hardening

Implements backpressure, validation, and queue management to stabilize ingestion under high load.

Install

mkdir -p .claude/skills/backend-telemetry-hardening && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/15150" && unzip -o skill.zip -d .claude/skills/backend-telemetry-hardening && rm skill.zip

Installs to .claude/skills/backend-telemetry-hardening

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Harden backend telemetry ingestion under load (MQTT spikes, DB pool pressure, validation failures).
99 charsno explicit “when” trigger
Advanced

Key capabilities

  • Quantify input/error rates, queue depth, and p95 latency to identify bottlenecks
  • Implement bounded queues and explicit backpressure for ingress
  • Throttle and batch MQTT messages per topic
  • Tune TimescaleDB pool limits and batch writes
  • Validate data at the boundary and route failures to a dead-letter queue
  • Apply idempotency guards, circuit-breaking, and bounded retries

How it works

The skill profiles current system performance to identify bottlenecks, then applies strategies like bounded queues, topic-based throttling, database pool tuning, and validation at the boundary.

Inputs & outputs

You give it
Telemetry ingestion under load with MQTT spikes, DB pool pressure, or validation failures
You get back
Stable telemetry ingestion with controlled queue/pool exhaustion and quarantined validation failures

When to use backend-telemetry-hardening

  • Debug telemetry ingestion lag
  • Optimize database pool usage
  • Harden MQTT ingestion

About this skill

TRIGGERS

  • MQTT bursts causing lag
  • TimescaleDB pool saturation/wait times
  • Ingestion queue growth/timeouts
  • Zod validation error spikes
  • Persistence inconsistency

STEPS

  1. Profile: Quantify input/error rates, queue depth, p95 latency. ID bottleneck (ingress/validation/persistence).
  2. Ingress: Bounded queues|Explicit backpressure|Separate parse/persistence|NO blocking CPU/sync I/O in hot paths.
  3. MQTT: Per-topic throttle/batch|Coalesce safe telemetry|Idempotent reconnects.
  4. TimescaleDB: Measure pool util/wait|Tune limits|Batch writes|Parameterized SQL|Selective columns/deterministic order.
  5. Validation (Zod): Validate at boundary|Route failures to dead-letter + reason|Preserve observability|NEVER bypass validation.
  6. Resilience: Idempotency guards|Circuit-breaking/load-shedding|Bounded/jittered retries.
  7. Verify: Burst & steady-state tests|Check pool stability, latency, data integrity.
  8. CI Checks:
    • pnpm --filter @ice-truck/backend lint
    • pnpm --filter @ice-truck/backend type-check
    • pnpm --filter @ice-truck/backend test

EXIT CRITERIA

  • Stable ingestion under load
  • NO uncontrolled queue/pool exhaustion
  • Validation failures quarantined, ZERO corruption

REFS

docs/DEFINITION_OF_DONE.md | docs/API.md | docs/THREAT_MODEL.md | backend/src/services/ | backend/tests/ | tests/k6/

How it compares

This approach systematically hardens telemetry ingestion paths against various load-induced failures, unlike ad-hoc fixes for individual issues.

Compared to similar skills

backend-telemetry-hardening side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
backend-telemetry-hardening (this skill)02moNo flagsAdvanced
langsmith-fetch67moReviewIntermediate
langfuse-common-errors11moCautionBeginner
autotel02moReviewIntermediate

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

langsmith-fetch

ComposioHQ

Debug LangChain and LangGraph agents by fetching execution traces from LangSmith Studio. Use when debugging agent behavior, investigating errors, analyzing tool calls, checking memory operations, or examining agent performance. Automatically fetches recent traces and analyzes execution patterns. Requires langsmith-fetch CLI installed.

67

langfuse-common-errors

jeremylongshore

Diagnose and fix common Langfuse errors and exceptions. Use when encountering Langfuse errors, debugging missing traces, or troubleshooting integration issues. Trigger with phrases like "langfuse error", "fix langfuse", "langfuse not working", "debug langfuse", "traces not appearing".

11

autotel

jagreehal

Use when instrumenting with trace/span/track, reviewing code for logging and observability patterns, converting console.log to wide events, adding structured errors, setting up canonical log lines, configuring init(), adding subscribers, or working in the autotel monorepo.

00

typescript-node-esm-compiler-runtime

GonkaGate

Own TypeScript plus Node.js ESM compiler/runtime correctness. Use whenever the real question is why TypeScript compiles but Node fails, how `tsconfig`/`package.json`/entrypoint/runtime mode must align, whether relative imports should use `.js` or `.ts`, how `nodenext`/`node20`/`verbatimModuleSyntax`

00

ai-debug-harness

DeandreFu

Self-driving Electron + Node debug harness with NDJSON logs, Playwright E2E, doctor preflight, and a YAML-frontmatter cookbook of known causes. Use when a voice-chat E2E flow misbehaves locally, audio publish silently fails, the agent worker emits warnings, or any LiveKit/Electron/Fastify error appe

00

debug-cross-service-auth

J-Akiru5

Use when: diagnosing Supabase JWT handoff failures from Next.js frontend to Laravel backend including headers, CORS, issuer, audience, and secret mismatches.

00

Search skills

Search the agent skills registry