boost-harness

To debug silent failures in an agent skill system, you analyze the architecture to detect design flaws like pipeline scatter and environment dependencies. This skill identifies structural weaknesses and facilitates guided, interactive improvements to reduce silent errors.

The best way to reduce token waste is by auditing the organization, sequencing, and validation approaches of your multi-skill system. This skill evaluates skill system architectures to find structural weaknesses and implements targeted improvements for efficiency.

You analyze agent skill architectures by running an automated audit on your skill directory to evaluate structural organization. This process detects common design smells like pipeline scatter and suggests interactive remediation for improved reliability.

Yes, you can automatically fix design smells by running an audit that suggests improvements and applies them after user approval. This interactive remediation targets structural weaknesses to prevent token waste and inconsistency.

Inconsistent outputs often result from structural weaknesses like pipeline scatter or missing layered guardrails in your skill system. Analyzing the architecture's validation approaches helps detect these design flaws and implement targeted reliability improvements.