agentsclimarketplace

Review me

Skill ShinewineW/lucideye/skills/review-me

A clearer second gaze for specs and artifacts: adversarial review skills for Claude Code.

Install
npx -y skills add ShinewineW/lucideye --skill review-me

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 14 days oldThe repository was created 14 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Multi-angle adversarial review of any artifact (spec, plan, ADR, or implementation). Use when the user wants a review, approval check, sign-off, or readiness assessment.

SKILL.md

1.4 KB, 253 tokens by cl100k_base, as published. Nobody here has run it

Review this artifact from multiple adversarial angles of your choosing — pick the lenses that matter most for THIS document.

SPEC mode: interrogate goals, boundaries, trade-offs, missing decisions. IMPL mode: interrogate mutations, migrations, error paths, observability.

Each angle must surface at least one finding — "no issues" means you haven't looked hard enough. When multiple angles flag the same issue, promote its severity. Do not ask questions; derive from context.

End with:

  1. Findings by angle, tagged [BLOCKER / CONCERN / NOTE]
  2. Clash — where angles disagree
  3. Verdict: approve | approve-with-conditions | request-changes | reject-for-now
  4. If approving: the single biggest risk you're accepting.

Enhanced Review

For high-stakes artifacts or reviews that warrant additional scrutiny:

  • Read references/dual-review.md and apply independent dual review, a both-pass verdict gate, fresh-reviewer fix cycles, and bounded batch sampling where appropriate.
  • Read references/review-orchestration.md and classify findings through RECONCILE, stop after a bounded review loop, and offer the documented cross-model escalation protocol when appropriate.

What ships with it: 3 files

18.8 KB alongside SKILL.md

agents/

Gives 0 of the 12 instructions most plan spec skills give in 253 tokens

Counted across 1,099 of the 1,860 authors here whose files we hold, read 2026-08-07

  • Ask one question at a timein 51 of 1099
  • Break plans into vertical slicesin 29 of 1099, across 11 files
  • Publish issues in dependency orderin 27 of 1099, across 9 files
  • Iterate until user approves the breakdownin 25 of 1099, across 7 files
  • Explore the repository to understand the codebase statein 24 of 1099, across 7 files
  • Use domain glossary vocabularyin 23 of 1099, across 5 files
  • Apply correct triage labels to published issuesin 23 of 1099, across 5 files
  • Prefer AFK slices over HITLin 22 of 1099, across 7 files
  • Write a specification before writing any codein 22 of 1099, across 14 files
  • Write failing tests before implementation codein 22 of 1099, across 20 files
  • Ask clarifying questions until requirements are concretein 21 of 1099, across 13 files
  • Respect existing architecture decision recordsin 20 of 1099, across 5 files

Said here and by no other author read

  • review the artifact from multiple adversarial angles
  • pick the review lenses that matter most for the document
  • interrogate goals boundaries and trade-offs for specs
  • interrogate mutations migrations and error paths for implementations
  • surface at least one finding per angle
  • promote severity when multiple angles flag the same issue

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.