How does the 'pressure test' feature work?

When a documentation fix is proposed, the skill runs batch tests via subagents against real historical failures to ensure the new instructions actually prevent the problematic behavior.

Does this skill require any external tools?

Yes, this skill is inoperable without the contextd MCP server, which provides the memory and ReasoningBank infrastructure required for analysis.

Can it predict future failures?

Yes, it uses pattern-based prediction to alert you when a current task matches conditions (like high complexity or specific keywords) that historically led to errors or skipped steps.

What is the primary purpose of the Self-Reflection skill?

It analyzes past agent interactions to identify systemic behavior patterns and provides pressure-tested improvements to your project's instructions and documentation.

Self-Reflection Agent Optimizer

Name: Self-Reflection Agent Optimizer
Author: fyrsmithlabs

byfyrsmithlabs

0•

Analytics & Monitoring

Analyzes Claude's behavior patterns to remediate documentation and optimize agent performance through deep memory analysis.

The Self-Reflection skill for contextd empowers Claude to perform recursive analysis on its own operational history. By mining memories and behavioral data, it identifies systemic patterns such as rationalized skips, overclaiming, and ignored instructions. It goes beyond simple error logging by performing causal chain analysis to find root causes in project documentation (CLAUDE.md) and provides pressure-tested remediations. This skill is essential for maintaining a healthy ReasoningBank and ensuring that agent behavior remains aligned with project requirements over long-term development cycles.

Key Features

01Causal Chain Analysis for tracing symptoms to root-cause documentation gaps

02Pressure-tested remediation flow using subagents to verify documentation fixes

030 GitHub stars

04ReasoningBank health monitoring for memory quality and tag hygiene

05Behavioral Taxonomy Reporting to categorize and prioritize agent failures

06Predictive risk alerts for high-complexity or security-adjacent tasks

Use Cases

01Conducting periodic audits of agent behavior to prevent instruction drift

02Improving CLAUDE.md after a series of failures or poor project outcomes

03Predicting and mitigating risks before starting complex, multi-step code refactors

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add fyrsmithlabs/marketplace self-reflection

For use in Claude.ai and ChatGPT

Key Features

01Causal Chain Analysis for tracing symptoms to root-cause documentation gaps

02Pressure-tested remediation flow using subagents to verify documentation fixes

030 GitHub stars

04ReasoningBank health monitoring for memory quality and tag hygiene

05Behavioral Taxonomy Reporting to categorize and prioritize agent failures

06Predictive risk alerts for high-complexity or security-adjacent tasks

Use Cases

01Conducting periodic audits of agent behavior to prevent instruction drift

02Improving CLAUDE.md after a series of failures or poor project outcomes

03Predicting and mitigating risks before starting complex, multi-step code refactors

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add fyrsmithlabs/marketplace self-reflection

For use in Claude.ai and ChatGPT