Skip to content

ByteTerrace/Puck / .claude/skills/review-passes

review-passes

Verified

Runs a Puck review pass, findings-only or review-and-fix. Use for review briefs and scope ranges, Codex task flags (--write, --cwd), job status, findings, blockers, follow-up fix branches.

FileKindSize
SKILL.mdskill md23.2 KB
evals/r01-brief-for-a-codex-pass/graders/contract-in-the-codes-names.mdmarkdown356 B
evals/r01-brief-for-a-codex-pass/graders/evidence-gathered.mdmarkdown234 B
evals/r01-brief-for-a-codex-pass/graders/exact-range-each-commit.mdmarkdown238 B
evals/r01-brief-for-a-codex-pass/graders/frame-names-the-domains.mdmarkdown119 B
evals/r01-brief-for-a-codex-pass/graders/hunt-per-area.mdmarkdown130 B
evals/r01-brief-for-a-codex-pass/graders/prompt-file.mdmarkdown72 B
evals/r01-brief-for-a-codex-pass/graders/rules-and-instructions-blocks.mdmarkdown264 B
evals/r01-brief-for-a-codex-pass/graders/skill-fired.mdmarkdown36 B
evals/r01-brief-for-a-codex-pass/graders/worktree-and-no-commit.mdmarkdown161 B
evals/r01-brief-for-a-codex-pass/prompt.mdmarkdown1.0 KB
evals/r02-landing-a-codex-pass/graders/checks-findings-against-files.mdmarkdown125 B
evals/r02-landing-a-codex-pass/graders/follow-up-branch.mdmarkdown217 B
evals/r02-landing-a-codex-pass/graders/no-co-authored-by.mdmarkdown150 B
evals/r02-landing-a-codex-pass/graders/no-second-round.mdmarkdown144 B
evals/r02-landing-a-codex-pass/graders/push-after-gates.mdmarkdown246 B
evals/r02-landing-a-codex-pass/graders/skill-fired.mdmarkdown36 B
evals/r02-landing-a-codex-pass/graders/wip-commit-unbuilt.mdmarkdown93 B
evals/r02-landing-a-codex-pass/prompt.mdmarkdown228 B
evals/r03-who-reviews-and-when-it-ends/graders/claude-agent-runs-the-pass.mdmarkdown200 B
evals/r03-who-reviews-and-when-it-ends/graders/lead-rules-the-rest.mdmarkdown171 B
evals/r03-who-reviews-and-when-it-ends/graders/no-second-round.mdmarkdown107 B
evals/r03-who-reviews-and-when-it-ends/graders/non-blocking-never-extends.mdmarkdown112 B
evals/r03-who-reviews-and-when-it-ends/graders/skill-fired.mdmarkdown36 B
evals/r03-who-reviews-and-when-it-ends/prompt.mdmarkdown317 B
evals/r04-pass-leaves-blockers-open/graders/lead-rules-each-blocker.mdmarkdown150 B
evals/r04-pass-leaves-blockers-open/graders/no-second-round.mdmarkdown73 B
evals/r04-pass-leaves-blockers-open/graders/report-to-the-lead.mdmarkdown122 B
evals/r04-pass-leaves-blockers-open/graders/scoped-fix-is-verified-not-reviewed.mdmarkdown126 B
evals/r04-pass-leaves-blockers-open/graders/skill-fired.mdmarkdown36 B
evals/r04-pass-leaves-blockers-open/prompt.mdmarkdown219 B
evals/r05-fix-pass-introduces-a-regression/graders/no-extra-round-for-a-verified-repair.mdmarkdown121 B
evals/r05-fix-pass-introduces-a-regression/graders/silent-results-is-a-hunt-class.mdmarkdown149 B
evals/r05-fix-pass-introduces-a-regression/graders/skill-fired.mdmarkdown36 B
evals/r05-fix-pass-introduces-a-regression/graders/verifier-repairs.mdmarkdown136 B
evals/r05-fix-pass-introduces-a-regression/prompt.mdmarkdown296 B
evals/r06-reviewer-claims-absent-from-the-tree/graders/checks-the-current-files.mdmarkdown118 B
evals/r06-reviewer-claims-absent-from-the-tree/graders/checks-the-pass-tree.mdmarkdown130 B
evals/r06-reviewer-claims-absent-from-the-tree/graders/dismisses-or-fixes.mdmarkdown133 B
evals/r06-reviewer-claims-absent-from-the-tree/graders/finding-is-evidence-not-a-verdict.mdmarkdown85 B
evals/r06-reviewer-claims-absent-from-the-tree/graders/skill-fired.mdmarkdown36 B
evals/r06-reviewer-claims-absent-from-the-tree/prompt.mdmarkdown224 B
evals/r07-write-pass-cannot-build/graders/brief-says-so.mdmarkdown100 B
evals/r07-write-pass-cannot-build/graders/no-restore.mdmarkdown69 B
evals/r07-write-pass-cannot-build/graders/restore-and-build-first.mdmarkdown133 B
evals/r07-write-pass-cannot-build/graders/sandbox-has-no-network.mdmarkdown121 B
evals/r07-write-pass-cannot-build/graders/skill-fired.mdmarkdown36 B
evals/r07-write-pass-cannot-build/prompt.mdmarkdown209 B
evals/r08-companion-job-keyed-by-cwd/graders/does-not-relaunch.mdmarkdown129 B
evals/r08-companion-job-keyed-by-cwd/graders/job-state-keyed-by-cwd-spelling.mdmarkdown128 B
evals/r08-companion-job-keyed-by-cwd/graders/same-spelling.mdmarkdown154 B
evals/r08-companion-job-keyed-by-cwd/graders/skill-fired.mdmarkdown36 B
evals/r08-companion-job-keyed-by-cwd/prompt.mdmarkdown270 B
evals/r09-range-given-as-two-dots/graders/base-as-a-commit.mdmarkdown109 B
evals/r09-range-given-as-two-dots/graders/folds-in-landings.mdmarkdown137 B
evals/r09-range-given-as-two-dots/graders/merge-base.mdmarkdown67 B
evals/r09-range-given-as-two-dots/graders/skill-fired.mdmarkdown36 B
evals/r09-range-given-as-two-dots/prompt.mdmarkdown243 B
evals/r10-task-help/graders/help-goes-to-the-model.mdmarkdown114 B
evals/r10-task-help/graders/names-the-flags.mdmarkdown108 B
evals/r10-task-help/graders/prompt-file-not-argv.mdmarkdown124 B
evals/r10-task-help/graders/skill-fired.mdmarkdown36 B
evals/r10-task-help/prompt.mdmarkdown195 B
evals/r11-broker-holds-the-worktree/graders/broker-outlives-the-job.mdmarkdown133 B
evals/r11-broker-holds-the-worktree/graders/skill-fired.mdmarkdown36 B
evals/r11-broker-holds-the-worktree/graders/stop-the-broker.mdmarkdown111 B
evals/r11-broker-holds-the-worktree/graders/then-remove.mdmarkdown101 B
evals/r11-broker-holds-the-worktree/prompt.mdmarkdown240 B
evals/r12-nonblocking-finding-after-the-pass/graders/no-another-pass.mdmarkdown116 B
evals/r12-nonblocking-finding-after-the-pass/graders/report-them.mdmarkdown87 B
evals/r12-nonblocking-finding-after-the-pass/graders/skill-fired.mdmarkdown36 B
evals/r12-nonblocking-finding-after-the-pass/graders/verify-the-fix-without-a-round.mdmarkdown108 B
evals/r12-nonblocking-finding-after-the-pass/prompt.mdmarkdown203 B
evals/r13-hard-doc-only-fixes-in-the-pass/graders/build-for-the-xml-change.mdmarkdown97 B
evals/r13-hard-doc-only-fixes-in-the-pass/graders/doc-checks-no-law.mdmarkdown171 B
evals/r13-hard-doc-only-fixes-in-the-pass/graders/lead-reads-the-doc-diff.mdmarkdown122 B
evals/r13-hard-doc-only-fixes-in-the-pass/graders/no-second-pass-for-doc-fixes.mdmarkdown138 B
evals/r13-hard-doc-only-fixes-in-the-pass/graders/skill-fired.mdmarkdown36 B
evals/r13-hard-doc-only-fixes-in-the-pass/prompt.mdmarkdown251 B
evals/r14-hard-author-moved-on-during-a-pass/graders/author-branch-untouched.mdmarkdown151 B
evals/r14-hard-author-moved-on-during-a-pass/graders/author-never-paused.mdmarkdown146 B
evals/r14-hard-author-moved-on-during-a-pass/graders/integration-lane-merges-both.mdmarkdown164 B
evals/r14-hard-author-moved-on-during-a-pass/graders/never-reset-or-rebase.mdmarkdown94 B
evals/r14-hard-author-moved-on-during-a-pass/graders/skill-fired.mdmarkdown36 B
evals/r14-hard-author-moved-on-during-a-pass/prompt.mdmarkdown241 B
evals/r15-hard-pass-adds-a-compat-alias/graders/cite-agents-rule-five.mdmarkdown98 B
evals/r15-hard-pass-adds-a-compat-alias/graders/dismiss-under-rule-five.mdmarkdown170 B
evals/r15-hard-pass-adds-a-compat-alias/graders/finding-is-evidence.mdmarkdown127 B
evals/r15-hard-pass-adds-a-compat-alias/graders/remove-the-alias.mdmarkdown86 B
evals/r15-hard-pass-adds-a-compat-alias/graders/skill-fired.mdmarkdown36 B
evals/r15-hard-pass-adds-a-compat-alias/prompt.mdmarkdown258 B
evals/r16-hard-pass-changed-nothing/graders/go-to-hand-over.mdmarkdown215 B
evals/r16-hard-pass-changed-nothing/graders/no-empty-wip.mdmarkdown72 B
evals/r16-hard-pass-changed-nothing/graders/skill-fired.mdmarkdown36 B
evals/r16-hard-pass-changed-nothing/graders/skip-the-verification-steps.mdmarkdown134 B
evals/r16-hard-pass-changed-nothing/prompt.mdmarkdown207 B
evals/r17-hard-verifier-rejects-a-fix/graders/law-that-cannot-fail.mdmarkdown147 B
evals/r17-hard-verifier-rejects-a-fix/graders/no-second-round.mdmarkdown106 B
evals/r17-hard-verifier-rejects-a-fix/graders/rejected-fix-is-not-landed.mdmarkdown90 B
evals/r17-hard-verifier-rejects-a-fix/graders/skill-fired.mdmarkdown36 B
evals/r17-hard-verifier-rejects-a-fix/graders/verifier-reports-rejected-with-reason.mdmarkdown133 B
evals/r17-hard-verifier-rejects-a-fix/prompt.mdmarkdown251 B
evals/r18-scope-two-dot-log-versus-diff/graders/checks-ancestry.mdmarkdown83 B
evals/r18-scope-two-dot-log-versus-diff/graders/lists-each-commit.mdmarkdown154 B
evals/r18-scope-two-dot-log-versus-diff/graders/log-versus-diff.mdmarkdown206 B
evals/r18-scope-two-dot-log-versus-diff/graders/pins-the-base.mdmarkdown162 B
evals/r18-scope-two-dot-log-versus-diff/graders/skill-fired.mdmarkdown36 B
evals/r18-scope-two-dot-log-versus-diff/prompt.mdmarkdown247 B
evals/r20-pass-ends-on-a-question/graders/answers-from-the-brief-and-rules.mdmarkdown167 B
evals/r20-pass-ends-on-a-question/graders/pass-is-not-done.mdmarkdown100 B
evals/r20-pass-ends-on-a-question/graders/prompt-file.mdmarkdown80 B
evals/r20-pass-ends-on-a-question/graders/resume-last.mdmarkdown70 B
evals/r20-pass-ends-on-a-question/graders/skill-fired.mdmarkdown36 B
evals/r20-pass-ends-on-a-question/graders/tighten-future-briefs.mdmarkdown142 B
evals/r20-pass-ends-on-a-question/prompt.mdmarkdown278 B
evals/r21-write-pass-while-a-run-builds/graders/mid-run-edit-changes-the-source.mdmarkdown174 B
evals/r21-write-pass-while-a-run-builds/graders/not-into-the-active-tree.mdmarkdown157 B
evals/r21-write-pass-while-a-run-builds/graders/skill-fired.mdmarkdown36 B
evals/r21-write-pass-while-a-run-builds/graders/wait-or-separate-worktree.mdmarkdown145 B
evals/r21-write-pass-while-a-run-builds/prompt.mdmarkdown337 B
evals/r23-review-never-blocks-landing/graders/no-review-for-an-ordinary-verb.mdmarkdown280 B
evals/r23-review-never-blocks-landing/graders/passes-fix-what-they-find.mdmarkdown281 B
evals/r23-review-never-blocks-landing/graders/review-never-blocks-landing.mdmarkdown218 B
evals/r23-review-never-blocks-landing/graders/skill-fired.mdmarkdown36 B
evals/r23-review-never-blocks-landing/prompt.mdmarkdown377 B
evals/r24-claude-pass-in-an-isolated-worktree/graders/detach-at-lane-head.mdmarkdown85 B
evals/r24-claude-pass-in-an-isolated-worktree/graders/follow-up-branch.mdmarkdown250 B
evals/r24-claude-pass-in-an-isolated-worktree/graders/own-worktree-not-integration.mdmarkdown206 B
evals/r24-claude-pass-in-an-isolated-worktree/graders/skill-fired.mdmarkdown36 B
evals/r24-claude-pass-in-an-isolated-worktree/prompt.mdmarkdown396 B
evals/r25-launch-command-for-a-sol-pass/graders/effort-high.mdmarkdown77 B
evals/r25-launch-command-for-a-sol-pass/graders/prompt-file-absolute.mdmarkdown104 B
evals/r25-launch-command-for-a-sol-pass/graders/skill-fired.mdmarkdown36 B
evals/r25-launch-command-for-a-sol-pass/graders/sol-model-id.mdmarkdown84 B
evals/r25-launch-command-for-a-sol-pass/graders/status-wait.mdmarkdown94 B
evals/r25-launch-command-for-a-sol-pass/graders/task-write.mdmarkdown82 B
evals/r25-launch-command-for-a-sol-pass/prompt.mdmarkdown349 B
evals/r26-reviewer-in-the-lead-checkout/graders/not-the-working-tree.mdmarkdown735 B
evals/r26-reviewer-in-the-lead-checkout/graders/object-commands.mdmarkdown77 B
evals/r26-reviewer-in-the-lead-checkout/graders/skill-fired.mdmarkdown36 B
evals/r26-reviewer-in-the-lead-checkout/prompt.mdmarkdown796 B
evals/r27-review-brief-without-revisions/graders/blocked-without-revisions.mdmarkdown748 B
evals/r27-review-brief-without-revisions/graders/skill-fired.mdmarkdown36 B
evals/r27-review-brief-without-revisions/prompt.mdmarkdown573 B
evals/r28-findings-only-pass-suspects-a-race/graders/null-check-reported.mdmarkdown489 B
evals/r28-findings-only-pass-suspects-a-race/graders/race-reported-unsettled.mdmarkdown730 B
evals/r28-findings-only-pass-suspects-a-race/graders/skill-fired.mdmarkdown36 B
evals/r28-findings-only-pass-suspects-a-race/prompt.mdmarkdown669 B
evals/r29-hunt-plan-for-a-cache-change/graders/hazards-by-name.mdmarkdown781 B
evals/r29-hunt-plan-for-a-cache-change/graders/reads-around-the-sites.mdmarkdown509 B
evals/r29-hunt-plan-for-a-cache-change/graders/skill-fired.mdmarkdown36 B
evals/r29-hunt-plan-for-a-cache-change/prompt.mdmarkdown656 B
evals/r30-findings-without-a-write-tool/graders/set-content.mdmarkdown68 B
evals/r30-findings-without-a-write-tool/graders/skill-fired.mdmarkdown36 B
evals/r30-findings-without-a-write-tool/graders/writes-the-file.mdmarkdown614 B
evals/r30-findings-without-a-write-tool/prompt.mdmarkdown704 B

evals/r17-hard-verifier-rejects-a-fix/prompt.md

View on GitHub
---
max_turns: 8
allowed_tools: [Read, Glob, Grep, Skill]
---

In the Puck repository, the Claude verifier rejects one of the Codex pass's four fixes as wrong, although that fix's new law passes. What happens to the finding, and is another pass owed?
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 323kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 323kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 323kUpdated Claude CodeCodexVerifiedInstall