autosearch:experience-capture

Append JSON execution events to per-skill patterns.jsonl files.

40|6|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/0xmariowu/Autosearch --skill autosearch-experience-capture
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: autosearch:experience-capture
Source: https://github.com/0xmariowu/Autosearch/tree/main/autosearch/skills/meta/experience-capture
Command: npx skills add https://github.com/0xmariowu/Autosearch --skill autosearch-experience-capture

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Leaf skills in AutoSearch generate events, and this meta-skill logs a single execution event to the per-skill experience/patterns.jsonl file. This enables downstream compaction into experience.md digest and supports pattern mining across skill runs.

Core Features & Use Cases

  • Append-only per-skill execution event to patterns.jsonl
  • Monthly archival and rotation of patterns.jsonl for long-term analysis
  • Feeds the experience-compact workflow that builds the experience.md digest from raw events

Quick Start

Call capture_event after each leaf-skill execution to append a JSON line to the skill's patterns.jsonl.

Frequently Asked Questions about autosearch:experience-capture

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I log skill execution events to a local file for pattern mining?

You can log skill execution events by calling the capture_event function after each run, which appends a JSON line to the skill's patterns.jsonl file. This creates an append-only record for pattern mining.

What is a patterns.jsonl file used for in skill telemetry?

A patterns.jsonl file stores append-only JSON execution events for each skill run. It serves as raw telemetry input for downstream compaction into an experience.md digest to support pattern discovery.

Do I need network access or an LLM to capture skill execution logs?

Capturing skill execution logs requires no network access or LLMs. The process operates with fast, local writes to append execution events directly to the per-skill patterns.jsonl file.

How do I archive and rotate execution logs for long-term analysis?

Execution logs support monthly archival and rotation of the patterns.jsonl file. This manages file growth over time while preserving the raw event data needed for long-term pattern analysis.

How does logged telemetry get compacted into an experience digest?

Logged telemetry in patterns.jsonl feeds the experience-compact workflow, which compacts raw events into an experience.md digest. This digest summarizes execution patterns for downstream analytics.