Tessl • The package manager for agent skills

Blog Docs Log inGet started

SkillPublic

model-routing

9-tier model routing system with cascading classifier fallback and result auto-evaluation

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/model-routing/SKILL.md

Quality

Content

67%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 body is a dense, concrete reference with executable commands, clear fallback workflows, and good section organization. Its main weaknesses are time-sensitive version/pricing data inline, internal inconsistencies (9 tiers claimed vs 8 listed; scripts referenced but not documented), and slight duplication between the pipeline diagram and fallback table.

Suggestions

Reconcile the '9-tier' claim with the routing table (only tiers 1-8 are listed) and document or remove the undocumented scripts ('~/bin/gemini-api', 'gemini-cli', 'grok') referenced in the Missing Wrapper Fallback section.

Move volatile model versions and per-million pricing into a separate reference file or a clearly marked 'current pricing' section so the core skill stays stable.

Trim the duplication between the ASCII pipeline diagram and the Classifier Fallback Chain table, and add a brief checkpoint for what to do when a delegation command fails (e.g. fall back to the next tier or handle directly).

Dimension	Reasoning	Score
Conciseness	Mostly efficient tables and commands, but the ASCII pipeline diagram duplicates the Classifier Fallback Chain table, and time-sensitive data (model versions like 'Gemini 2.5 Flash-Lite', per-million prices, 'Kimi K2.6') is inline with no deprecated/old-patterns section, which the guidelines penalize.	3 / 5
Actionability	Provides mostly executable guidance — exact commands ('~/bin/gemini --flash-lite "prompt"'), env vars, exit-code contract — but has minor gaps: the '9-Tier Routing Table' lists only tiers 1-8, and the missing-wrapper section cites scripts ('~/bin/gemini-api', 'gemini-cli', 'grok') absent from the Available Scripts listing.	4 / 5
Workflow Clarity	The pipeline diagram gives a clear sequence with explicit failure branches, cascading fallback levels, and missing-wrapper handling, but there is no explicit checkpoint for what to do when a delegation command itself fails mid-task (retry, re-route, or handle directly).	4 / 5
Progressive Disclosure	A well-organized single-file reference with clear section headers and cohesive inline tables; no bundle files exist to reference. It exceeds 50 lines so the simple-skill exception does not apply, and the detailed routing table and Maggy integration could arguably live in separate reference files.	4 / 5
	Total	15 / 20 Passed

Description

53%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 states a clear, distinctive 'what' with a couple of concrete mechanisms, but it lacks any 'use when' trigger guidance and relies on technical jargon rather than natural trigger phrases. Adding an explicit when-to-use clause with user-facing trigger terms would raise both completeness and trigger term quality.

Suggestions

Append a 'Use when...' clause to the description with concrete trigger phrases, e.g. 'Use when configuring or debugging prompt classification, model delegation, or routing prompts to the cheapest capable model.'

Include natural trigger terms users would actually say (e.g. 'model cost', 'cheap model', 'delegate to a model', 'classifier failing') alongside the technical terms.

Name one or two concrete capabilities (e.g. 'routes prompts across 9 cost tiers via ~/bin delegation scripts with cached-tier recovery') to make the 'what' more actionable.

Dimension	Reasoning	Score
Specificity	Names the domain ('model routing system') and two concrete mechanisms ('cascading classifier fallback', 'result auto-evaluation'), but lists no concrete actions or capabilities, matching the 'domain and 1-2 concrete actions, not comprehensive' anchor.	3 / 5
Completeness	The 'what' is clear (9-tier routing with fallback and auto-evaluation) but the description field itself contains no 'Use when...' clause or equivalent trigger guidance, which caps completeness at 3 per the judging guidelines.	3 / 5
Trigger Term Quality	Contains relevant keywords ('model routing', 'classifier', 'fallback') but they skew technical and miss natural user phrasings like 'cheap model', 'cost', or 'delegate', fitting 'some relevant keywords but missing common variations'.	3 / 5
Distinctiveness Conflict Risk	'9-tier model routing system' carves a clear niche unlikely to fire for unrelated skills, but the technical phrasing leaves minor overlap risk with other orchestration or cost-optimization skills, matching 'mostly distinct; minor overlap risk'.	4 / 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: alinaqi/maggy
Path: skills/model-routing/SKILL.md
Revision: 55aabe2

Reviewed: 1 day 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.

model-routing

Table of Contents

Is this your skill?