Tessl • The management layer for AI agent skills

Blog Docs Log inGet started

SkillPublic

judge-with-debate

Evaluate solutions through multi-round debate between independent judges until consensus

Quality

67%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

High

Do not use without reviewing

Fix and improve this skill with Tessl

tessl review fix ./plugins/sadd/skills/judge-with-debate/SKILL.md

Quality

Content

62%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 skill has a clear, well-sequenced workflow with strong actionability via concrete templates and dispatch blocks, but it is hampered by verbosity from structural redundancy and conversational filler, and it inlines everything in a single file rather than progressively disclosing detail into reference files.

Suggestions

Remove the duplicated orchestration walkthrough (the 'Orchestration Instructions' section restates the Process phases) and delete conversational filler like 'Let's work through this systematically'.

Extract the large meta-judge and judge prompt templates into separate reference files (e.g. references/meta-judge-prompt.md, references/judge-prompt.md) and link to them, reducing SKILL.md to an overview.

Tighten the consensus-check guidance into a compact checklist rather than a step-by-step prose explanation.

Dimension	Reasoning	Score
Conciseness	The ~470-line body is noticeably verbose: orchestration steps are described twice (Process phases then again under 'Orchestration Instructions') and it contains conversational filler ('Let's work through this systematically', 'Let's synthesize the evaluation results step by step', 'Let's compile the final consensus') that pads without adding guidance.	2 / 5
Actionability	Provides concrete, mostly-copy-paste-ready prompt templates, explicit Task tool dispatch blocks with model and subagent_type, file naming conventions, and a worked example with actual scores — executable guidance with only minor placeholder gaps.	4 / 5
Workflow Clarity	Clear multi-phase sequence (Phase 0/0.5/1/2/3) with explicit consensus-checking logic, decision points (consensus achieved / next round / report no consensus), and a feedback loop of up to 3 debate rounds with re-validation each round.	5 / 5
Progressive Disclosure	No bundle files exist and all content — large prompt templates and the entire orchestration logic — is inlined in one long SKILL.md; section headers give some structure, but content that would belong in separate reference files is inline with no external references.	3 / 5
	Total	14 / 20 Passed

Description

57%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 clearly conveys what the skill does — multi-round debate-based evaluation — but omits an explicit 'Use when' trigger clause, which caps its completeness. Trigger term coverage is good though not exhaustive.

Suggestions

Add an explicit 'Use when...' clause naming concrete triggering scenarios, e.g. 'Use when you need rigorous, multi-perspective evaluation of a solution and single-pass review is insufficient'.

Broaden trigger terms to include natural synonyms users say, such as 'review', 'score', 'assess', or 'critique'.

Mention 1-2 more concrete actions (e.g. 'generate tailored rubrics', 'challenge ratings with evidence') to raise specificity.

Dimension	Reasoning	Score
Specificity	Names the domain (solution evaluation) and a concrete action ('multi-round debate between independent judges') but does not enumerate multiple specific actions, matching the anchor for 1-2 concrete actions without comprehensive coverage.	3 / 5
Completeness	The description clearly states what the skill does but lacks any 'Use when...' trigger clause; per the judging guidelines a missing explicit trigger guidance caps completeness at 3.	3 / 5
Trigger Term Quality	Includes natural phrases users might say ('Evaluate solutions', 'multi-round debate', 'independent judges', 'until consensus') with good coverage, though synonyms like 'review', 'score', or 'assess' are missing.	4 / 5
Distinctiveness Conflict Risk	The multi-round debate evaluation niche is somewhat specific, but evaluation/review skills broadly overlap, so it could still overlap with similar skills.	3 / 5
	Total	13 / 20 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: NeoLabHQ/context-engineering-kit
Path: plugins/sadd/skills/judge-with-debate/SKILL.md
Revision: 23e2428

Reviewed: 14 days ago

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.

judge-with-debate

Table of Contents

Is this your skill?