auto-review-loop-minimax
Autonomous multi-round research review loop using MiniMax API. Use when you want to use MiniMax instead of Codex MCP for external review. Trigger with "auto review loop minimax" or "minimax review".
64
76%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/skills-codex/auto-review-loop-minimax/SKILL.mdQuality
Discovery
89%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 is strong on completeness, trigger terms, and distinctiveness, clearly specifying when and why to use this skill with explicit trigger phrases and differentiation from Codex MCP. Its main weakness is a lack of specificity about what concrete actions the 'research review loop' actually performs—what does it review, what outputs does it produce, what does 'multi-round' entail?
Suggestions
Add specific concrete actions describing what the review loop does, e.g., 'Sends code or documents to MiniMax API for iterative feedback, collects review comments, and applies suggested changes across multiple rounds.'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | It names the domain ('research review loop') and mentions the MiniMax API, but doesn't list specific concrete actions like what the review entails, what outputs are produced, or what steps the loop performs. | 2 / 3 |
Completeness | Clearly answers both 'what' (autonomous multi-round research review loop using MiniMax API) and 'when' (use when wanting MiniMax instead of Codex MCP for external review, with explicit trigger phrases). | 3 / 3 |
Trigger Term Quality | Includes explicit trigger phrases ('auto review loop minimax', 'minimax review') and natural keywords like 'MiniMax API', 'external review', and 'Codex MCP' that users would naturally use when requesting this skill. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive due to the specific mention of MiniMax API, explicit contrast with Codex MCP, and unique trigger phrases. Very unlikely to conflict with other skills. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
62%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill excels at actionability and workflow clarity — it provides a thorough, well-sequenced autonomous review loop with concrete commands, state recovery, and explicit stop conditions. However, it suffers significantly from verbosity: the same API call patterns (both MCP and curl) are repeated three times across different sections, roughly tripling the token cost without adding information. Consolidating the duplicated prompt templates into a single reference section would dramatically improve token efficiency.
Suggestions
Consolidate the API call examples: define the MCP and curl patterns once (in API Configuration), then reference them in Phase A and the Round 2+ section instead of repeating the full code blocks.
Move the Round 2+ prompt template into a separate referenced file (e.g., PROMPT_TEMPLATES.md) since it's a large block that duplicates structure already shown in Phase A.
Remove the explanatory note about why MiniMax is used instead of Codex — this is background context that doesn't help Claude execute the skill and wastes tokens.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose with massive duplication. The curl fallback and MCP examples are repeated three times (API Configuration, Phase A, and Prompt Template sections), nearly tripling the token cost. The round 2+ prompt template is shown in full for both methods, duplicating what was already shown in Phase A. Much of this could be consolidated. | 1 / 3 |
Actionability | The skill provides fully executable curl commands, concrete MCP tool invocations, specific JSON schemas for state files, exact file paths, and detailed prompt templates. Everything is copy-paste ready with clear parameter placeholders. | 3 / 3 |
Workflow Clarity | The workflow is exceptionally well-structured with clear phases (A through E), explicit stop conditions with precise criteria (score >= 6 AND verdict matching), initialization logic handling fresh start vs resume vs stale state, validation checkpoints, and error recovery via state persistence. The feedback loop (review → fix → re-review) is the core design. | 3 / 3 |
Progressive Disclosure | The skill references shared protocols at the bottom with clear links, which is good. However, the body itself is monolithic (~250+ lines) with significant inline duplication that could be factored into separate reference files (e.g., prompt templates, API configuration). The curl examples repeated three times are a clear case of content that should be referenced rather than inlined. | 2 / 3 |
Total | 9 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Commit
31cc7b9
- Reviewed
Table of Contents
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.