optimize

Drive structured autoresearch iteration after evo:discover and the baseline commit. Use when the user invokes /evo:optimize or asks to try ideas, try variants, run experiments, use available GPUs, improve the current best/frontier, continue an evo search, or compare candidate changes in an evo workspace. The orchestrator plans and spawns optimization subagents; candidate edits/runs belong to those subagents. Width is set via subagents=N (1 for serial workloads, larger for parallel); the loop's structural value applies at any width.

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

81%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.

A thorough, highly actionable orchestration protocol with clear sequencing and validation checkpoints, though it runs long and repeats several invariants. External references are well signaled and one level deep, but the body itself could be split further.

Suggestions

Consolidate the repeated resource-cap and candidate-delegation invariants into a single stated invariant, referencing it from the loop steps instead of restating.

Move the prose-vs-workflow driver reconciliation and the full per-host spawn-command table into a dedicated reference file, keeping only the decision rule inline.

Extract the per-host enumeration in 'Host conventions' since the same host-specific spawn shapes are detailed again in Step 5, leaving one canonical location.

Dimension	Reasoning	Score
Conciseness	Mostly efficient domain protocol Claude would not already know, but noticeably padded: the resource-cap and candidate-delegation invariants are restated, the prose-vs-workflow driver and its reconciliation are explained at length, and per-host spawn shapes are enumerated both in 'Host conventions' and again in Step 5.	3 / 5
Actionability	Fully executable guidance throughout — concrete copy-paste CLI commands (evo scratchpad, evo frontier, evo wait --for ..., evo prune --exhausted), a verbatim subagent prompt, and exact jq/bash blocks covering the common cases.	5 / 5
Workflow Clarity	The Loop is explicitly sequenced (Step 0 → 8) with validation checkpoints (read .evo/project.md first, 'verify every pattern before emitting it', diversity check, 'check yourself' subagent-count check) and feedback loops (GATE_FAILED → fix → re-run; evo run recovery), satisfying the batch-operation validation requirement.	5 / 5
Progressive Disclosure	Clear section headers and well-signaled one-level-deep references to sibling-skill docs (sizing-the-round.md, evo-wait.md, cli-quick-reference.md, provider-matrix.md), but the body is a long near-monolithic wall with inlined detail (workflow-vs-prose reconciliation, per-host enumeration) that could live in a separate reference.	4 / 5
	Total	17 / 20 Passed

Description

95%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.

A strong, third-person description that clearly states what the skill does and when to use it, with rich natural-language triggers. It is slightly procedural in its action list rather than comprehensively enumerating capabilities, but otherwise excellent.

Dimension	Reasoning	Score
Specificity	Names several concrete actions ('plans and spawns optimization subagents; candidate edits/runs belong to those subagents. Width is set via subagents=N') plus compare candidate changes, but is procedural rather than a comprehensive action list, leaving minor gaps.	4 / 5
Completeness	Explicitly answers both what ('Drive structured autoresearch iteration... orchestrator plans and spawns optimization subagents') and when ('Use when the user invokes /evo:optimize or asks to try ideas...') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Comprehensive natural trigger coverage a user would actually say — 'try ideas', 'try variants', 'run experiments', 'use available GPUs', 'improve the current best/frontier', 'continue an evo search', 'compare candidate changes' — plus the /evo:optimize slash command.	5 / 5
Distinctiveness Conflict Risk	Highly specific to a post-discover evo workspace optimization loop with its own /evo:optimize trigger; clear niche with minimal overlap risk against other skills.	5 / 5
	Total	19 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (592 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
referenced_paths_exist	Referenced path issues: 2 missing	Warning

	Total	13 / 16 Passed

Repository: evo-hq/evo
Path: plugins/evo/skills/optimize/SKILL.md
Commit: ab5fbd6

Reviewed: 5 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.