optimize

Use when an algorithm-first quest should manage candidate briefs, optimization frontier, branch promotion, or fusion-aware search instead of the paper-oriented default loop.

Quality

85%

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

Quality

Content

85%

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

The content is well-structured and highly actionable with a clear workflow and strong progressive disclosure into real reference files, but it loses conciseness to visible line-level duplication that should be de-duplicated.

Suggestions

Remove the duplicated/near-duplicate bullets and lowercase restatements (the two Git-branch lines, the repeated 'recommend one approach...', 'only skip smoke...', 'gate evolution on clear objective signal', and the repeated 'Do not treat one candidate creation or one smoke pass as stage completion') to tighten conciseness.

Fold the empty '### reference-file' appendix stubs into the existing 'Operational guidance' reference list so there is one navigation surface instead of two, slightly redundant ones.

Consolidate the 'Constraints' three-object-levels list with the 'Core object model' section, which restates the same candidate-brief / durable-line / implementation-attempt taxonomy.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient domain guidance, but it is padded with duplicated lines (e.g. consecutive 'Do not create a new Git branch/worktree...' bullets and repeated lowercase restatements like 'recommend one approach...' / 'only skip smoke...' / 'gate evolution on clear objective signal'), so it could be tightened.	2 / 3
Actionability	It gives concrete, specific guidance — named artifact calls (artifact.get_optimization_frontier, artifact.submit_idea with submission_mode, artifact.record with report_type, bash_exec, memory.search), six named submodes, a default selection order, and explicit route meanings — which is actionable for an instruction-only skill.	3 / 3
Workflow Clarity	The 5-step Control workflow is clearly sequenced, backed by a Validation checklist, a mandatory first-call sequence, and explicit Exit criteria, providing the checkpoints the rubric rewards.	3 / 3
Progressive Disclosure	SKILL.md stays an overview and delegates detail to 13 real one-level-deep references under 'Operational guidance' with a clear when-to-use list; all referenced paths exist in references/.	3 / 3
	Total	11 / 12 Passed

Description

85%

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 is complete and distinctive, with concrete domain actions and an explicit 'Use when' trigger, but its trigger terms lean on internal jargon rather than natural user phrasing.

Suggestions

Add one or two plain-language trigger terms a user might actually say (e.g. 'tune', 'search-loop', 'optimize a method') alongside the domain jargon to improve trigger_term_quality.

Consider naming the single most common user-facing scenario first so the trigger reads naturally before the internal vocabulary.

Dimension	Reasoning	Score
Specificity	Lists several concrete domain actions — 'manage candidate briefs, optimization frontier, branch promotion, or fusion-aware search' — rather than vague language, matching the 'multiple specific concrete actions' anchor.	3 / 3
Completeness	It explicitly answers 'when' via the 'Use when an algorithm-first quest should manage...' clause and answers 'what' through the named management activities and the contrast 'instead of the paper-oriented default loop'.	3 / 3
Trigger Term Quality	Terms like 'algorithm-first quest', 'fusion-aware search', and 'branch promotion' are system-internal jargon a user would not naturally say, so coverage of natural trigger words is partial rather than strong.	2 / 3
Distinctiveness Conflict Risk	The narrow 'algorithm-first quest' niche and the explicit 'instead of the paper-oriented default loop' contrast carve out a clear, non-overlapping trigger space.	3 / 3
	Total	11 / 12 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: ResearAI/DeepScientist
Path: src/skills/optimize/SKILL.md
Commit: b366244

Reviewed: 2 days 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.