model-right-sizer-dryrun

Preview the model-routing MAP for an intent without building anything. You give it a free-text INTENT — "build a CLI that…", "add a feature that…", "refactor X into Y" — and it invokes the `model-right-sizer` agent in BLUEPRINT-ONLY mode: decompose the work, score each piece, and emit a single schema-conformant JSON blueprint (task→model→effort→budget→ schema→confidence, per `schemas/blueprint.schema.json`), then STOP. No build, no file edits, no after-the-fact usage report — it is the what-would-this-cost / how-would-this-route preview lever, safe to run against any idea, and the JSON it returns is what an orchestrator parses to route dispatch. Read-only. Use when someone says "dry-run the right-sizer on …", "what's the map for …", "how would you route …", or "show me the blueprint for … before I build it".

Quality

89%

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

85%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, well-sequenced body with a concrete validation command and error-recovery loop. The only weakness is conciseness: the conceptual framing and the repeated agent/install relationship prose add padding that could be tightened without losing clarity.

Suggestions

Trim or fold the 'Why blueprint-only is not a new mode' section into a single sentence in the intro; the Pass A/Pass B distinction is already implied by the 'STOP' constraint.

Remove the duplicated model-right-sizer-install relationship prose from the intro — it is already covered in the Related section — to cut tokens.

Consider tightening the validator-rationale paragraph (step 4) to the concrete command plus one line on why eyeballing fails, rather than the multi-sentence justification.

Dimension	Reasoning	Score
Conciseness	Mostly efficient, but the 'Why blueprint-only is not a new mode' conceptual framing, the intro restating the description, and the mandate/install relationship (explained in both the intro and the Related section) are unnecessary explanation that could be trimmed.	3 / 5
Actionability	Fully executable guidance: a copy-paste validator command ('echo "$BLUEPRINT_JSON" \| uv run --no-project --with jsonschema scripts/validate_blueprint.py -'), explicit agent dispatch params (mode: "dry_run", intent), and a concrete re-emit-on-failure loop that quotes validator errors.	5 / 5
Workflow Clarity	Clear numbered 1–4 sequence with an explicit validation checkpoint, an error-recovery feedback loop (non-zero exit → ask agent to re-emit once → re-validate), and an explicit STOP; read-only, so the destructive-cap does not apply.	5 / 5
Progressive Disclosure	Well-organized sections (What to do, What this does NOT do, Related) with clearly signaled one-level-deep external references (schemas, agent markdown, sibling skill) and no nested reference chains.	5 / 5
	Total	18 / 20 Passed

Description

95%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, specific description with explicit what/when coverage and natural trigger phrases. The only real weakness is second-person voice ('You give it a free-text INTENT'), which costs it a specificity point under the voice guideline.

Suggestions

Rewrite the description in third person (e.g., 'Previews the model-routing MAP for an intent… Given a free-text intent, it invokes…') to satisfy the voice guideline and recover the specificity point.

Dimension	Reasoning	Score
Specificity	Names multiple concrete actions ('decompose the work, score each piece, and emit a single schema-conformant JSON blueprint (task→model→effort→budget→schema→confidence)'), but the second-person phrasing 'You give it a free-text INTENT' triggers the voice penalty, dropping it from 5.	4 / 5
Completeness	Explicitly answers both 'what' (preview the model-routing MAP, emit a schema-conformant JSON blueprint, then STOP; no build/edits/report) and 'when' (a 'Use when…' clause with concrete trigger phrases).	5 / 5
Trigger Term Quality	Comprehensive natural trigger phrases a user would actually say: 'dry-run the right-sizer on …', 'what's the map for …', 'how would you route …', and 'show me the blueprint for … before I build it'.	5 / 5
Distinctiveness Conflict Risk	Occupies a clear 'dry-run / preview / blueprint-before-building' niche with distinctive triggers, separable from the build agent and the install sibling.	5 / 5
	Total	19 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 13 suspicious	Warning
referenced_paths_exist	Referenced path issues: 2 missing	Warning

	Total	13 / 16 Passed

Repository: Cloudzero/cloudzero-claude-marketplace
Path: plugins/model-right-sizer/skills/model-right-sizer-dryrun/SKILL.md
Commit: 62c27f3

Reviewed: 1 day 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.