Name: roast
Rating: 68 (1 reviews)
Author: alirezarezvani

roast

Use when someone asks to roast an idea, pressure-test or stress-test an idea, validate a business idea, "convene the panel", get a brutal second opinion before building something, or says "/roast". Spins up a 5-angle panel (Critic, Champion, Analyst, Investigator, Customer) that attacks the idea from every angle, then a Judge returns one GO / RESHAPE / KILL verdict with the cheapest test to de-risk it.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

The canonical home for this skill is roast in alirezarezvani/claude-skills

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.

A well-structured, highly actionable operational skill body with concrete prompts, commands, and output templates, weakened by modest duplication and, critically, a completely missing bundle: every referenced file and script is absent, so the script-driven workflow cannot run and the References/Assets sections are dead links.

Suggestions

Ship the referenced bundle files — scripts/brief_builder.py, scripts/verdict_synthesizer.py, scripts/cheapest_test_designer.py, the three references/*.md, and both assets/*.md — or remove the dead References, Assets, and Tooling sections; as delivered, every referenced path is broken and the workflow's script steps cannot execute.

Enumerate the flag combinations for scripts/cheapest_test_designer.py across all six named risk types (demand/price/feasibility/differentiation/channel/retention); only '--risk price --price 99' is shown.

Trim duplication: drop or shrink the Tooling table (it restates the scripts' roles from Steps 1–3) and the 'Distinct From' section (it repeats the frontmatter metadata.distinct_from).

Dimension	Reasoning	Score
Conciseness	Mostly lean and operational — verbatim panelist prompts, exact commands, and a fixed output template all earn their tokens. Minor duplication: the Tooling table restates the three scripts' roles already explained in Steps 1–3, 'Distinct From' repeats the frontmatter metadata, and 'Rules'/'Anti-Patterns' overlap on averaging and softening.	4 / 5
Actionability	Copy-paste-ready bash commands with example values, five verbatim panelist prompts, and an exact verdict output shape make the guidance highly concrete. Minor gaps: the cheapest_test_designer example shows only one risk/flag combination, and every command depends on script files that are absent from the bundle, so they cannot actually execute as delivered.	4 / 5
Workflow Clarity	A clear three-step sequence with an explicit pre-panel validation gate (brief_builder 'tells you if anything critical is still missing before you spend five subagents') and a deterministic synthesizer step with a hard 'do not average' rule. Minor gaps: no recovery guidance if a panelist returns off-format output or a script invocation fails.	4 / 5
Progressive Disclosure	The body is well-sectioned and references are clearly signaled one level deep with descriptions, but the bundle contains no references/, scripts/, or assets/ directories — all seven referenced paths (3 references, 3 scripts, 2 assets) are dangling, and the three load-bearing scripts the workflow's commands invoke do not exist. Scored against the actual (empty) bundle structure, the disclosure chain is broken rather than merely imperfect.	2 / 5
	Total	14 / 20 Passed

Description

100%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 is exemplary: third-person voice, comprehensive natural trigger terms including the slash command, an explicit what-and-when structure, and concrete named capabilities (five personas, GO/RESHAPE/KILL verdict, cheapest de-risking test) with no vague padding.

Dimension	Reasoning	Score
Specificity	'Spins up a 5-angle panel (Critic, Champion, Analyst, Investigator, Customer) that attacks the idea from every angle, then a Judge returns one GO / RESHAPE / KILL verdict with the cheapest test to de-risk it' names multiple specific concrete actions covering the full workflow. Not the 4 anchor: there are no coverage gaps — panel composition, verdict form, and test output are all explicit.	5 / 5
Completeness	Explicitly answers both: 'Use when someone asks to roast an idea... or says "/roast"' (when) and 'Spins up a 5-angle panel... returns one GO / RESHAPE / KILL verdict with the cheapest test' (what), with concrete trigger phrases. Not the 4 anchor: the 'when' clause is fully explicit with multiple concrete triggers, not merely adequate.	5 / 5
Trigger Term Quality	'roast an idea, pressure-test or stress-test an idea, validate a business idea, "convene the panel", get a brutal second opinion before building something, or says "/roast"' comprehensively covers natural phrasings and synonyms a user would actually say. Not the 4 anchor: no common variation of the request is missing.	5 / 5
Distinctiveness Conflict Risk	A clear niche — adversarial 5-persona idea panel with a decisive verdict — with distinct trigger phrases ('roast', 'pressure-test', 'convene the panel') that are unlikely to fire for adjacent skills. Not the 4 anchor: overlap risk is minimal, not minor-but-present.	5 / 5
	Total	20 / 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: 5 missing	Warning
referenced_paths_exist	Referenced path issues: 17 missing	Warning

	Total	13 / 16 Passed

Repository: alirezarezvani/claude-skills
Path: .gemini/skills/roast/SKILL.md
Commit: 19392f7

Reviewed: 1 hour 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.