ablation-planner

Use when main results pass result-to-claim (claim_supported=yes or partial) and ablation studies are needed for paper submission.

Quality

67%

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/ablation-planner/SKILL.md

Quality

Content

75%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 content is a well-structured, actionable workflow that respects Claude's intelligence and includes real validation checkpoints for a batch operation; its main gap is the absence of an explicit error-recovery loop for failed runs.

Dimension	Reasoning	Score
Conciseness	The body is efficient and assumes Claude's ML competence — no padding explaining what ablations are — with only minor redundancy between the Codex prompt's requested fields and the parsed-plan output template.	4 / 5
Actionability	It provides a concrete Codex MCP call with model, config, and a full prompt template plus specific file targets (EXPERIMENT_LOG.md, research_contract.md) and naming conventions; minor gaps remain where steps like "Create configs/scripts for each ablation" stay high-level.	4 / 5
Workflow Clarity	Five steps are clearly sequenced with a feasibility gate (Step 4), a smoke-test checkpoint (Step 5.2), and a budget feedback loop with Codex; it falls just short of 5 because no error-recovery path is given for failed smoke tests or crashed ablation runs.	4 / 5
Progressive Disclosure	No bundle files exist, but the single-file body is well-organized into clear sections (When to Use, Workflow Steps 1–5, Rules) with easy navigation; it is appropriately self-contained with only minor organization room.	4 / 5
	Total	16 / 20 Passed

Description

60%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 has an explicit, distinctive trigger and good natural keywords, but it weakly conveys the skill's action — "what" is implied by the name rather than stated — which leaves completeness and specificity below their top anchors.

Suggestions

Lead with an explicit action verb for the "what", e.g. "Designs and plans ablation studies...", so the capability is stated directly rather than implied by the skill name.

Add a few natural synonyms a reviewer might say (e.g. "ablation experiments", "sensitivity studies") to round out trigger-term coverage.

Keep the strong "Use when..." clause but pair it with a one-phrase "what" so both what and when are explicitly answered.

Dimension	Reasoning	Score
Specificity	The description names the domain ("ablation studies are needed for paper submission") but states no concrete action verb — it only expresses that ablations are needed, not what the skill does (design/plan them), which is merely implied by the skill name.	2 / 5
Completeness	The "when" is explicitly and specifically answered ("Use when main results pass result-to-claim... and ablation studies are needed for paper submission"), but the "what" is only weakly implied rather than stated as an action, so it does not fully reach the both-answered anchor of 4.	3 / 5
Trigger Term Quality	Natural domain terms an ML researcher would say are present ("ablation studies", "paper submission", "claim_supported=yes or partial", "result-to-claim"), giving good coverage with only a few common variants missing.	4 / 5
Distinctiveness Conflict Risk	The trigger is tied to a specific project workflow condition (claim_supported=yes or partial after result-to-claim), giving it a clear niche with minimal overlap risk against other skills.	5 / 5
	Total	14 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Path: skills/ablation-planner/SKILL.md
Commit: 0472e53

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.