reflecting

Review completed tasks, verify quality gates, and generate reflection artifacts.

11|1|Updated Apr 10, 2026

Search Tags:#quality assurance #security review #reflection #backlog generation #workflow evolution #tests verification #operating-model gate

Ready to run

npx skills add https://github.com/Arcanada-one/datarim --skill reflecting

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: reflecting
Source: https://github.com/Arcanada-one/datarim/tree/main/skills/reflecting
Command: npx skills add https://github.com/Arcanada-one/datarim --skill reflecting

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It reduces inconsistent project outcomes by forcing a structured review after every completed task and turning lessons learned into actionable improvements.

Core Features & Use Cases

Definition of Done review: Validates that completion criteria were truly met.
Test, security, and reflection artifacts: Verifies tests pass, checks for security vulnerabilities, and writes a reflection document to capture what happened.
Class A/B self-evolution proposals with safety gates: Proposes improvements, routes Class A for approval, blocks risky operating-model changes (Class B) until the PRD/contract is updated, and enforces a stack-agnostic gate plus regression-proofing via bats.

Quick Start

Ask your AI operator to run the archive reflection step by completing a task with Datarim so Step 0.5 generates the reflection and follow-up backlog items automatically.

Frequently Asked Questions about reflecting

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

FAQPage Schema

How do I automate quality assurance and reflection after completing a project task?▼

Automated quality assurance and reflection are triggered by running the archive reflection step. This validates completion criteria, verifies tests, checks for security vulnerabilities, and generates reflection documents with backlog suggestions automatically.

How does workflow evolution handle risky operating-model changes?▼

Workflow evolution enforces safety gates by blocking risky operating-model changes (Class B) until the PRD or contract is updated. Class A improvement proposals are routed for approval and regression-proofed via bats verification.

What is the best way to generate a backlog from completed work across project phases?▼

Generating a backlog from completed work requires reading the reviewer persona and resolving the datarim path safely. A runtime template then produces reflection artifacts and follow-up tasks across requirements, design, execution, QA, compliance, and archival phases.

Can I enforce stack-agnostic gating and tests verification for Class A changes?▼

Yes, you can enforce stack-agnostic gating and tests verification for Class A changes. The process applies regression-proofing via bats verification to ensure safe workflow evolution before returning follow-up tasks to the archive Step 4 consumer.

When do I need to update the PRD or contract for proposed self-evolution improvements?▼

You must update the PRD or contract when proposed self-evolution improvements involve risky operating-model changes. These Class B changes are explicitly blocked by safety gates until the PRD or contract is updated to reflect the new workflow.

What is the definition of done review process for iterative project execution?▼

The definition of done review process validates that completion criteria were truly met during iterative project execution. It captures what happened by writing reflection artifacts, verifying tests pass, and checking for security vulnerabilities before archiving.