Find the first useful log clue before the noise takes over.
Ignition failures rarely arrive as one neat error. This skill teaches the AI to preserve timestamps, source labels, logger names, caps, and time windows so it can separate the first useful signal from the follow-on chatter.
Why it was made
- Keeps Gateway logs, wrapper logs, browser console output, audit rows, alarm journal rows, and diagnostic bundles separated by source.
- Uses narrow time windows, logger/message filters, severity caps, and truncation metadata before reading more.
- Turns logs into next evidence requests instead of guessing a product bug from one stack trace.
Common use cases
- Perspective errors after a page deploy.
- Gateway startup, module load, or service wrapper failures.
- Noisy logs where the first useful error is buried under repeated warnings.
- Support handoffs that need sanitized evidence, not raw log dumps.
Proof style
Validate the artifact, not the promise.
The skill prefers current vendor docs or release notes before saying an error is a known product issue. Logs explain what happened; they do not prove page rendering or production root cause by themselves.
Evidence
Log triage workflow proof
The graphic is AI-generated for the site. The checks below come from sanitized log-assistant evidence: runner health, Gateway context, rejected broad misuse, focused log queries, wrapper rotation handling, and timeline/next-evidence synthesis. Logs are diagnostic clues, not proof by themselves that a page rendered or that a vendor bug exists.
- health confirmed logQuery plus non-primary and rotated wrapper-log support before any log reads.
- A missing-filter logQuery was rejected with a structured 400 response instead of becoming an unbounded scan.
- Focused queries preserved source files, line-scan counts, matched-before-cap counts, truncation flags, and tail-window state.
- A narrow follow-up query returned 12 source-labeled ERROR rows after the broader exception sweep was capped.
- Timeline and next-evidence runs kept capped windows, source labels, and missing human context attached to the handoff.
Example prompt
Use the Ignition Log Diagnostics Helper to triage these Gateway and wrapper logs from a failed Perspective deploy. Identify the first useful error, separate repeated noise, and tell me the next evidence to collect.This is the kind of short instruction the skill is designed to make useful. The value is in the checks and Ignition-specific defaults the AI uses after reading the skill.
Expected output
- Labels each source, time window, severity, logger, and cap.
- Separates the first suspicious event from repeated follow-on messages.
- Reports likely cause, confidence, missing evidence, and the next focused query.
Related workflows
Skills that pair with this one
Lifetime membership
Want access to this skill?
The one-time membership unlocks this skill, the full Ignition AI Toolkit, the Web Dev runner materials, setup docs, versioned downloads, and future Ignition 8.3 skills as they ship.
No subscription. Lifetime members keep portal access to the current toolkit and future skill releases, including planned Ignition 8.3-focused skills as they are added.