skills.rest
nickgogan
nickgoganCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

identify-artifacts

Classify research findings into pattern, skill, rule, template, or agent forms using a rubric.

Updated Apr 19, 2026
Search Tags:#report generation#knowledge base#subagents#classification#rubric#triage#research findings
One-click install
npx skills add https://github.com/nickgogan/improvement-loop --skill identify-artifacts-nickgogan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: identify-artifacts
Source: https://github.com/nickgogan/improvement-loop/tree/main/.claude/skills/identify-artifacts
Command: npx skills add https://github.com/nickgogan/improvement-loop --skill identify-artifacts-nickgogan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When a knowledge base accumulates many research findings, deciding which artifact form each finding should become (pattern, skill, rule, template, or agent) is slow and inconsistent. This Skill automates that classification with a calibrated rubric and produces a reviewable identification report. ## Core Features & Use Cases - Rubric-based classification: Applies a form-classification rubric mechanically via parallel Sonnet subagents, assigning exactly one form per finding with confidence levels and tier dispatch (auto/guided/hitl). - Filtering and dedup: Skips weak-evidence, already-adopted, and already-extracted findings before classification. - Structured report output: Writes an identification report with summary tables, per-finding rationale, priority-revision proposals, and guide-cluster routing checks that downstream skills consume. - Use Case: After a research loop extracts 20 new findings, run this Skill to classify them all, review the HITL items, then hand approved findings to the extraction step. ## Quick Start Run the identify-artifacts skill on all P1 findings to generate a classification report for review.

Frequently Asked Questions about identify-artifacts

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

FAQPage Schema
How do I classify research findings into artifact forms automatically?▼

Run this Skill with a priority filter (P1, P2) or specific finding names. It batches findings, dispatches Sonnet subagents with the classification rubric, and writes an identification report with assigned forms, confidence levels, and rationale.

What is the difference between a pattern, skill, rule, template, and agent artifact?▼

A pattern is a reusable design approach, a skill is an ordered procedure, a rule is a binary constraint at an enforcement boundary, a template is a fillable scaffold with variables, and an agent is a named role with a durable cognitive disposition. The rubric's exclusion criteria and level-of-abstraction discriminator resolve ambiguous cases.

Does this Skill draft or write the artifacts it classifies?▼

No. It performs classification only and writes an identification report. Drafting and writing artifacts is handled by the separate extract-artifacts step, which consumes the report's Status and Form fields after human review.

How are low-confidence or ambiguous classifications handled?▼

Findings with LOW confidence are routed to the HITL (human-in-the-loop) tier and flagged with the ambiguity in the report. MED confidence goes to a guided review tier, and HIGH confidence goes to an auto tier ready for extraction.

What happens to findings with weak evidence or duplicates?▼

Findings marked as weak anecdotal evidence or already adopted are filtered out before classification unless explicitly named. The Skill also scans the extracts directory to skip findings that already have staged artifacts, reporting the filtered counts.

Related Skills

Explore More Product & Management Skills→
brainbrewlabsbrainbrewlabs
7

analysis

Analyze qualitative research texts and output a structured markdown analysis report.

Official
Intermediate
maxculen86maxculen86

issue-creation

Automate GitHub issue creation with mandatory templates and automatic labeling.

Community
Intermediate
torikhayestorikhayes

security-findings-link

Link duplicate security-review findings so consolidated reports merge them into one entry.

Community
Basic
arbiterForgearbiterForge
139

security-architecture

Map attack surfaces and identify unmitigated STRIDE threats before implementation.

Official
Advanced