anneal

Use when the user wants to systematically fix AI code slop — duplicated logic, over-engineering, silent error swallowing, convention drift, cargo-cult patterns, and other LLM-introduced architectural decay — over a specified duration

Quality

82%

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

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 highly actionable, well-sequenced orchestration skill with explicit validation checkpoints and a complete subagent template. Its main cost is redundancy in the anti-early-exit material and the duplication between the dot graph and the numbered steps.

Suggestions

Drop either the dot graph or the 'Step by step' list — they encode the identical control flow, and the numbered list with exact commands is the more actionable of the two.

Consolidate 'Preventing Premature Exit', 'Preventing Sabotaged Runs', and 'Red Flags' into a single anti-drift section; the same 'the clock decides, dispatch again' message is currently repeated in three places plus a ten-row table.

Move the slop catalog to a references/slop-catalog.md file and pass subagents its path (or instruct them to read it); this keeps the orchestrator's dispatch-loop context lean while still delivering the full catalog to workers.

Dimension	Reasoning	Score
Conciseness	The body is imperative and assumes Claude's competence, but contains real redundancy: the dot graph restates the numbered steps, 'Red Flags' overlaps 'Preventing Sabotaged Runs', and 'Preventing Premature Exit' restates the Iron Law roughly ten ways across a table and three sections. This fits 'mostly efficient but could be tightened' rather than the minor-trimming of anchor 4.	3 / 5
Actionability	Fully executable guidance throughout: exact commands (`date +%s`), exact progress-file and summary templates, a complete subagent prompt template with commit-message format, and a concrete revert-on-failure protocol. Copy-paste ready for the common cases.	5 / 5
Workflow Clarity	Clear sequenced loop with explicit validation checkpoints: tests run before and after every change, revert mandated on new failures, time-check before every dispatch, and a stall-recovery feedback loop. This satisfies the batch/destructive validation requirement at the highest anchor.	5 / 5
Progressive Disclosure	With no bundle files, everything is inline in one well-sectioned document with clear headers and a quick-reference table — good structure and navigation. It falls short of anchor 5 only because the ~60-line slop catalog, which is handed to subagents anyway, is a natural candidate for a references/ file in a 280-line skill.	4 / 5
	Total	17 / 20 Passed

Description

83%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 description: it explicitly states what it does (fix enumerated AI-slop patterns) and when to use it, with distinctive niche trigger terms. Its only weaknesses are a single action verb and missing common synonyms like 'refactor' or 'tech debt'.

Dimension	Reasoning	Score
Specificity	Enumerates concrete problem types ('duplicated logic, over-engineering, silent error swallowing, convention drift, cargo-cult patterns') but relies on a single action verb ('fix'), so it lists several specific targets rather than the multiple distinct concrete actions of the anchor-5 example.	4 / 5
Completeness	Explicitly answers both what ('systematically fix AI code slop' with an enumerated pattern list) and when ('Use when the user wants to...') with concrete trigger phrases and the duration-input expectation, matching the anchor-5 example.	5 / 5
Trigger Term Quality	Explicit 'Use when the user wants to systematically fix AI code slop' with recognizable natural phrases ('AI code slop', 'duplicated logic', 'over-engineering'), but misses common synonyms like 'clean up', 'refactor', or 'tech debt', so coverage is good rather than comprehensive.	4 / 5
Distinctiveness Conflict Risk	'AI code slop', 'LLM-introduced architectural decay', and the duration-bound framing form a clear niche, but it could still trigger for general cleanup/refactoring requests that overlap with closely related skills.	4 / 5
	Total	17 / 20 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: av/harbor
Path: .agents/skills/anneal/SKILL.md
Commit: e38ae4c

Reviewed: 1 day 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.