Explore

Find agent skills by outcome

90,082 skills indexed with the new KISS metadata standard.

Showing 24 of 90,082Categories: Coding & Debugging, General, Openclaw, Communication

General

PromptBeginner5 minmarkdown

- Produce corrective actions that are specific

measurable

Mar 21, 2026

Coding & Debugging

PromptBeginner5 minmarkdown

RULE: When using this prompt

you must create a file named `TODO_rca.md`. This file must contain the findings resulting from this research as checkable checkboxes that can be coded and tracked by an LLM.,TRUE,TEXT,wkaandemir

Find agent skills by outcome

- Produce corrective actions that are specific

**RULE:** When using this prompt

Refactoring Expert Agent Role

- **Postmortem Quality**: Evaluate clarity

- Separate what is known from what is suspected

- [ ] File paths

- [ ] Analysis focuses on systems and controls

Before finalizing

Use checkboxes and stable IDs (e.g.

Use checkboxes and stable IDs (e.g.

- **Evidence**: Concrete logs

- **Ignoring Detection Gaps**: Focusing only on preventing the root cause while neglecting improvements to monitoring

In `TODO_rca.md`

- **Single-Cause Tunnel Vision**: Focusing on one causal factor while ignoring contributing conditions

- **Untestable Recommendations**: Proposing corrective actions without verification criteria

- Review circuit breaker states

- **Missing Evidence Trail**: Drawing conclusions without citing specific logs

- **Symptom-Level Conclusions**: Stopping the analysis at the immediate trigger (e.g.

- Apply structured log queries with timestamp ranges

- Use centralized logging (ELK Stack

- Identify log gaps caused by retention policies

- Leverage distributed tracing (Jaeger

- Use Prometheus

- Check APM tools for error rate spikes

RULE: When using this prompt

- Postmortem Quality: Evaluate clarity

- Evidence: Concrete logs

- Ignoring Detection Gaps: Focusing only on preventing the root cause while neglecting improvements to monitoring

- Single-Cause Tunnel Vision: Focusing on one causal factor while ignoring contributing conditions

- Untestable Recommendations: Proposing corrective actions without verification criteria

- Missing Evidence Trail: Drawing conclusions without citing specific logs

- Symptom-Level Conclusions: Stopping the analysis at the immediate trigger (e.g.