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

77%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, clearly sequenced skill with strong validation feedback loops and concrete templates. Its main weaknesses are motivational repetition across several guardrail sections and a lack of progressive disclosure — everything lives in one long file with no bundled references.

Suggestions

Consolidate the overlapping guardrail sections (Preventing Premature Exit, Preventing Sabotaged Runs, Stall Recovery, Red Flags) into a single concise checklist to remove repeated 'check the clock, dispatch' messaging.

Move the full Slop Catalog and the verbatim subagent prompt into reference files (e.g. references/slop-catalog.md, references/subagent-prompt.md) and point to them from SKILL.md, keeping the overview lean.

Tighten the slop catalog entries so each pattern's 'Look for' and 'Fix' guidance is one line, trusting Claude's existing knowledge of these anti-patterns.

Dimension	Reasoning	Score
Conciseness	Mostly efficient and free of basic-concept padding, but the 'check the clock and dispatch' message is restated across Preventing Premature Exit, Preventing Sabotaged Runs, Stall Recovery, and Red Flags sections, adding motivational repetition that could be tightened.	3 / 5
Actionability	Provides copy-paste-ready commands (`date +%s`), concrete file-path templates (`/tmp/anneal-<timestamp>.md`), a commit-message format, markdown templates for the progress file, and a verbatim subagent prompt covering the common case.	5 / 5
Workflow Clarity	A graphviz diagram plus numbered step-by-step instructions include explicit validation checkpoints (run tests before and after every change) and a clear revert-on-failure feedback loop for this batch/destructive operation.	5 / 5
Progressive Disclosure	Well-organized with clear headers and a quick-reference table, but it is a long monolithic single file with no references, and content that could live separately (the full 10-entry slop catalog, the subagent prompt) is inlined.	3 / 5
	Total	16 / 20 Passed

Description

87%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, distinctive description with an explicit Use-when trigger and a concrete, enumerated catalog of targets. The only weakness is that it relies on a single action verb ('fix') rather than listing multiple distinct actions.

Dimension	Reasoning	Score
Specificity	Names the domain and enumerates concrete targets ('duplicated logic, over-engineering, silent error swallowing, convention drift, cargo-cult patterns'), but the only action verb is 'fix' applied to that list rather than multiple distinct actions.	4 / 5
Completeness	Explicitly answers both 'what' (systematically fix AI code slop with enumerated patterns) and 'when' via a concrete 'Use when the user wants to...' trigger phrase including the duration condition.	5 / 5
Trigger Term Quality	Includes natural phrases a user might say ('AI code slop', 'duplicated logic', 'over-engineering', 'cargo-cult patterns') with good coverage, though common variations like 'clean up AI-generated code' or 'refactor LLM code' are absent.	4 / 5
Distinctiveness Conflict Risk	Carves a clear niche (AI-introduced code slop remediation over a specified duration) with distinctive triggers and minimal overlap risk against general refactoring or linting skills.	5 / 5
	Total	18 / 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: 67cbe66

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.