agent-orchestration-improve-agent

Systematic improvement of existing agents through performance analysis, prompt engineering, and continuous iteration.

1.54x

Quality

53%

Does it follow best practices?

Impact

82%

1.54x

Average score across 3 eval scenarios

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/agent-orchestration-improve-agent/SKILL.md

Quality

Content

57%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The skill is well-structured as a phased workflow with clear sequencing and good validation/rollback checkpoints, but it suffers from a monolithic single-file layout with no progressive disclosure and relies on placeholder pseudocode rather than executable examples. Tightening the verbosity and moving detailed material into referenced files would materially improve it.

Suggestions

Split the detailed metric templates and A/B config into references/ files (e.g. references/metrics-template.md, references/ab-testing.md) and link to them one level deep, leaving SKILL.md as an overview.

Replace placeholder pseudocode ([X%], [User request], [Reasoning]) with concrete runnable examples or clearly justified templates so guidance is copy-paste ready.

Trim the extended-thinking preamble and the duplicated description/intro line, and cut the lists of universally-known concepts (e.g. basic metric definitions) to reduce token cost.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient lists and commands, but it includes a heavy extended-thinking preamble and restates the description, with some sections (e.g. 3.3 Evaluation Metrics) enumerating concepts Claude already knows, so it could be tightened.	2 / 3
Actionability	It gives concrete invocation blocks ('Use: context-manager', 'Command: analyze-agent-performance $ARGUMENTS') and a few formulas, but many blocks are pseudocode templates with placeholders ([X%], [User request]) rather than executable, copy-paste-ready instructions.	2 / 3
Workflow Clarity	The four phases are clearly sequenced, each with numbered substeps, and Phase 4 includes explicit validation checkpoints and rollback triggers ('Switch to previous stable version', 'Fix and re-test before retry') for the risky deployment operations.	3 / 3
Progressive Disclosure	It is a monolithic ~340-line single file with no bundle files in references/scripts/assets and no one-level-deep references to split out the detailed templates; everything is inline in one wall of content.	1 / 3
	Total	8 / 12 Passed

Description

50%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description states a clear domain and high-level actions but stops short of concrete verbs or any explicit 'Use when' trigger guidance. It is competent but generic, leaving the when-to-use decision implicit.

Suggestions

Add an explicit 'Use when...' clause naming natural user phrases like 'improving an agent', 'agent keeps failing', or 'tune agent performance'.

Replace abstract nouns ('performance analysis', 'prompt engineering') with concrete actions ('run A/B tests on prompts', 'roll back regressions').

Name trigger variations users actually say (e.g. 'agent reliability', 'eval suite', 'failure analysis') to improve distinctiveness.

Dimension	Reasoning	Score
Specificity	Names a domain ('agent performance') and a few actions ('performance analysis, prompt engineering, continuous iteration') but these are abstract categories rather than concrete actions like 'A/B test prompts' or 'rollback failed changes'.	2 / 3
Completeness	It states what the skill does but provides no 'Use when...' clause or explicit trigger guidance, which per the rubric caps completeness at 2.	2 / 3
Trigger Term Quality	'Improving an existing agent's performance' is a phrase a user might say, but it lacks common natural variations like 'agent is failing', 'tune agent', 'eval suite', or 'rollback' that users would actually utter.	2 / 3
Distinctiveness Conflict Risk	Scoped to 'existing agents' and optimization rather than agent building, which is somewhat distinct, but the phrasing is generic enough to overlap with general debugging or prompt-writing skills.	2 / 3
	Total	8 / 12 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: rmyndharis/antigravity-skills
Path: skills/agent-orchestration-improve-agent/SKILL.md
Commit: e63f7dd

Reviewed: about 5 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.