Back to catalogue

Big-Log Extractor & Summariser

Developer Official v1.0.1

Extracts the signal from a large log deterministically — pull matching lines, dedupe, count — then has a model summarise only the small, unique result, cutting tokens dramatically.

by skrptiq

logs builtin observability
workflow Updated 19 July 2026 3 nodes

Sign in to import this skrpt into your workspace.

Sign in
What's included 3 nodes
1 workflow 1 skill 1 prompt 3 built-ins
workflow (1)

Big-Log Extractor & Summariser

Extract matching lines from a log, dedupe, count, then summarise the small unique set.

skill (1)

Summarise Errors

Summarises the small unique set of errors.

prompt (1)

Summarise Errors

Summarises the unique error set.

Built-ins (3) deterministic · runs offline · no tokens

Regex extract

built-in · offline · 0 tokens

Pull every match of a regular expression out of the text.

Sort lines

built-in · offline · 0 tokens

Sort the input lines, optionally reversed and/or de-duplicated.

Count

built-in · offline · 0 tokens

Count lines, words, characters, or JSON-array elements in the input.

Requirements 1 service
Services
  • LLM Service
Security Passed
All checks passed v1.0.1 · scanner v3.3.0

Detected

  • Services: llm-service
  • Permissions:
  • Data Handling:
Version history 1 release
v1.0.1 latest 12 July 2026

GH#820 — migrate the 3 builtin steps (extract-errors, dedupe-errors, count-unique) to the node-less `local.builtin` shape (#802 D1b): `id`/`input`/`context`, no backing skill nodes/edges; chained builtins take `input: {{steps.<prev-id>.output}}` (were from_step bindings); downstream refs `{{steps.Count Unique/Dedupe Errors.output}}` → `{{steps.count-unique/dedupe-errors.output}}`. Deterministic output unchanged. contents skills 4→1.

More from Developer

View all →

Send feedback

Sign in to send feedback — it lets us follow up and keeps the tracker free of spam.