agent-orchestration-improve-agent

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

Quality

49%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

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

48%Weight 40%Scale 1-5

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

The skill presents a well-sequenced, validation-aware optimization workflow, but it is verbose and padded with general AI concepts Claude already knows, and it inlines ~350 lines of methodology that should be split into reference files. Actionability is mixed: concrete thresholds and commands sit alongside pseudo-invocations and placeholder templates.

Suggestions

Cut conceptual explanations Claude already knows (what chain-of-thought, few-shot, Constitutional AI, A/B testing, and statistical significance are) and keep only the skill-specific application and commands.

Move the detailed test-framework, prompt-engineering technique catalog, and metrics reference into separate reference files (e.g. TESTING.md, PROMPT_TECHNIQUES.md) and link to them from a lean SKILL.md overview.

Replace pseudo-invocations and bracketed templates ('Use: prompt-engineer / Technique: ...', [User request]) with concrete, copy-pasteable commands or fully worked examples.

Dimension	Reasoning	Score
Conciseness	The body is noticeably verbose and extensively explains concepts Claude already knows (chain-of-thought, few-shot learning, Constitutional AI, role definition, A/B testing methodology, statistical significance, semver). It is above anchor 1 only because some concrete commands and thresholds earn their place; it does not reach 3 due to the volume of padded conceptual explanation.	2 / 5
Actionability	It offers some concrete guidance (context-manager commands, parallel-test-runner config, rollback thresholds like 'success rate drops >10%', sample sizes, version format), but much is pseudocode-style agent invocations ('Use: prompt-engineer / Technique: ...') and bracketed template placeholders ([User request], [Why this works]). This is anchor 3: some concrete guidance but incomplete, rather than 4's mostly-executable bar.	3 / 5
Workflow Clarity	A clear four-phase sequence (Analysis → Prompt Engineering → Testing → Deployment) with explicit validation (Phase 3 test suite + A/B framework), rollback feedback loops (detect → alert → switch → analyze → fix → re-test), and checklists (success criteria, rollback triggers). It falls short of 5 because Phase 2 reads as a technique catalog rather than a tightly sequenced workflow with per-step checkpoints.	4 / 5
Progressive Disclosure	The document has good section structure, but it is a single ~350-line monolith with no references to separate files, and substantial methodology (test framework, prompt techniques, metrics reference) that belongs in reference files is inlined. It is above anchor 2 thanks to clear sectioning but below 4 because nothing is split out.	3 / 5
	Total	12 / 20 Passed

Description

50%Weight 40%Scale 1-5

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 clearly states what the skill does but omits any 'Use when' trigger guidance, capping completeness at 3. Its actions are methodological and jargon-leaning rather than concrete natural phrases.

Suggestions

Add an explicit 'Use when...' clause naming concrete triggers, e.g. 'Use when improving an existing agent's performance, analyzing agent failure modes, or running A/B tests on prompt changes.'

Replace generic process labels with more concrete actions (e.g. 'analyzes agent performance metrics, rewrites prompts, runs A/B test suites') to lift specificity.

Include natural synonyms users would actually say ('optimize my agent', 'improve agent reliability', 'agent not performing well') to broaden trigger coverage.

Dimension	Reasoning	Score
Specificity	Names the domain ('improvement of existing agents') and three methodological actions ('performance analysis, prompt engineering, and continuous iteration'), but they are generic process labels rather than concrete deliverables and coverage is not comprehensive. It sits at anchor 3 rather than 4 because the actions lack specificity and omit testing/rollout that the body covers.	3 / 5
Completeness	It gives a clear 'what' (systematic improvement of existing agents) but provides no 'when' / 'Use when' trigger guidance, which caps completeness at 3 per the rubric guideline. It is not a 2 because the 'what' is clear, and not a 4 because no explicit trigger clause exists.	3 / 5
Trigger Term Quality	Contains relevant keywords like 'improvement of existing agents' and 'performance' a user might say, but leans on jargon ('prompt engineering', 'continuous iteration') and misses common natural synonyms or phrasings. It is above anchor 2 (one or two generic keywords) but below 4 due to the jargon and absent synonyms.	3 / 5
Distinctiveness Conflict Risk	The 'improvement of existing agents' framing is somewhat specific, but it overlaps with other agent-orchestration skills (build, debug, evaluate) and lacks distinct triggers. It fits anchor 3 (somewhat specific but could overlap) rather than 4's 'mostly distinct' niche.	3 / 5
	Total	12 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: sickn33/antigravity-awesome-skills
Path: skills/agent-orchestration-improve-agent/SKILL.md
Commit: 1f67c44

Reviewed: about 16 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.