neb-constructive-analysis

Automate structured investigative analysis with evidence-backed findings across debugging, code reviews, and system assessments.

Updated Aug 23, 2026

Search Tags:#investigation #code-review #root-cause #incident-analysis #blame-free #evidence-backed

Ready to run

npx skills add https://github.com/nebinfra/nebcore-ai-skills --skill neb-constructive-analysis

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: neb-constructive-analysis
Source: https://github.com/nebinfra/nebcore-ai-skills/tree/main/neb-constructive-analysis
Command: npx skills add https://github.com/nebinfra/nebcore-ai-skills --skill neb-constructive-analysis

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Research-first investigation pipeline — enforces constructive, blame-free analysis with evidence-backed claims, enabling teams to diagnose issues, review code, and deliver actionable findings without blame.

Core Features & Use Cases

Six-phase investigation workflow: GATHER → ANALYZE → CONTEXTUALIZE → DRAFT → TONE-CHECK → DELIVER.
Activates when the prompt-investigation-gate injects it via additionalContext, and can be referenced directly by engineers seeking structured analysis.
Supports on-demand references and prior-memory integration through the references/ directory to enrich findings.

Quick Start

Feed the latest incident artifacts (logs, configs, code history) into the six-phase pipeline to generate a structured constructive-analysis draft.

Frequently Asked Questions about neb-constructive-analysis

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

FAQPage Schema

How do I conduct a blame-free root cause analysis for a software incident?▼

Blame-free root cause analysis is automated by feeding incident artifacts like logs and code history into a six-phase investigative pipeline that produces structured, evidence-backed findings without assigning blame.

What is the best way to structure code review findings with evidence?▼

Code review findings are structured using the Observation, Context, Recommendation (OCR) format, ensuring every claim is evidence-backed and contextualized before delivery.

How do I calibrate the depth of an incident investigation?▼

Incident investigation depth is calibrated during the drafting phase by selecting Quick Triage, Standard, or Deep Root Cause Analysis levels based on the complexity of the available artifacts.

Can I integrate prior memory and references into a system assessment?▼

Prior memory and references can be integrated into a system assessment by loading supplementary context from the references directory to enrich the investigative findings.

Does the constructive analysis workflow verify findings before delivery?▼

The constructive analysis workflow enforces a mandatory verification step during the Tone-Check and Deliver phases to ensure all claims are validated and constructively phrased before final output.

What do I need to start an evidence-backed investigation of a debugging incident?▼

To start an evidence-backed investigation, you need to gather incident artifacts such as application logs, system configurations, and code history to feed into the initial analysis phase.