Menu
Verify an implementer's diff with an INDEPENDENT, different-vendor sub-agent (diff plus contract only), LOCALLY and BEFORE the PR is opened; turn blocking issues into fix-tasks and loop until clean — only then does the implementer open its PR.
Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.
Baseado na classificação ocupacional SOC
Use when bootstrapping a repository — greenfield or brownfield — into Uhstray's standard development conventions, OpenSpec, and the two-store memory setup. Idempotent and non-destructive: safe to re-run, never overwrites existing docs. Triggers on "repo kickstart", "bootstrap this repo", "set up our conventions", "conform this repo", "initialize plans/openspec/memory", "initialize this repo for openspec and memory", "set up two-store memory here", "index this repo and create its memory bank".
Use when checking or setting up the four-strata memory system on this machine and in the current project — "is my memory set up right", "memory onboarding", "set up memory", "check my memory setup", after a fresh machine/repo, when repo-kickstart or buhhdy's preflight reports memory degradation, or when .claude/memory / MemPalace / evolve looks misconfigured.
Use when any orchestrating agent's development workflow — buhhdy's Workflow 2 is the canonical caller — has open PRs that need driving from creation through merge and cleanup — the terminal step once PRs are open. Triggers on "shepherd this PR", "drive the PR to merge", "babysit the PRs", "monitor the PR to merge", and post-merge close-out (close linked issues, archive the change, remove the worktree, janitor stale branches in the orchestrator's own prefix (buhhdy/* canonical)). Also load to verify PR-lifecycle preconditions (branch protection, CodeRabbit) before shepherding.
Use when the user says "remember", "don't forget", "save that", "keep that in mind", "recall", "what do we know about", "look that up", or any request to persist or retrieve project knowledge — stores memories in .claude/memory/ using the Claude Code memory format, committed to git so the whole team shares context.
Use when starting new planning/research on a project or picking up development against an existing plan — buhhdy's two standard, repeatable sequences: Planning & Research (fresh problem to a validated plan — an OpenSpec change where the repo has adopted the conventions — plus issues) and Development (fanout, audit, docs, and pr-shepherd).
Reference skill for buhhdy's provider routing logic — the decision tree, model tier table, per-provider strengths, skill-to-provider affinity, and cross-review pairings. Load when reasoning about which provider to use and why, or when a task doesn't clearly match the main routing tree.
| name | cross-review |
| description | Verify an implementer's diff with an INDEPENDENT, different-vendor sub-agent (diff plus contract only), LOCALLY and BEFORE the PR is opened; turn blocking issues into fix-tasks and loop until clean — only then does the implementer open its PR. |
The implementer never signs off on its own work — a different model does, and review is a sub-agent that returns a structured report, not a transcript anyone needs to read through.
Cross-review runs LOCALLY, BEFORE the PR is created (config.yaml's Cross-Review Rule, layer 1). The implementer resolves every blocking finding first; only a clean review earns the PR. CodeRabbit then reviews the PR independently after creation (layer 2 — never pre-empted or duplicated here), and pr-shepherd owns everything after PR creation (layer 3).
sys_os_shell("git -C .worktrees/<task_id> status --porcelain") must
return EMPTY. Staged, unstaged, or untracked files mean the branch diff
would miss work — send the task back to the implementer to commit
everything first; never review a dirty worktree. Then get the diff —
sys_os_shell("git -C .worktrees/<task_id> diff main...HEAD"). (Only
for a PR that already exists — e.g. re-reviewing a fix pushed to an
open PR under pr-shepherd — use gh pr diff <pr> instead.)sys_os_shell, executed FROM the task's worktree (e.g.
sys_os_shell("cd .worktrees/<task_id> && <test cmd>")) so they
validate the diff under review, never buhhdy's own checkout — gates run
anywhere else are false-green evidence. If red, re-dispatch the
implementer to drive it green first; don't involve the reviewer yet.codex, gemini-*, or opencode; Codex built it → claude_code,
gemini-*, or opencode; a gemini-* worker built it → claude_code,
codex, or opencode; opencode built it → claude_code, codex, or
gemini-*. The three gemini-* workers are ONE vendor and never review
each other; gemini-* defaults to gemini-standard, or gemini-complex
for large/multimodal diffs). Use a task-based title such as
review-auth-refactor, never the raw vendor name:
sys_session_send(agent="claude_code"|"codex"|"gemini-standard"|"gemini-complex"|"opencode", title="review-<task_slug>", args={purpose: "review", input: "<the diff> + <the acceptance contract>. Review ONLY against the contract. Report blocking / non-blocking / suggestions. Do not edit code."}). Give it the diff as text — do NOT point
it at the implementer's worktree. Fetch the diff and emit the
sys_session_send call in the SAME turn you decide to review — never end a
turn having only announced "I'll load cross-review and fetch the diff" with
no tool call (that dropped turn stalls the run; nothing dispatches and no
inbox wake arrives). Once the reviewer dispatch is in flight, end your turn;
collect the inbox-delivered structured report with sys_read_inbox when it
returns. Use sys_session_get_history only to debug an empty or unclear
review result.sys_session_send — reuse the original implementer's
agent + title (or address it by session_id) with
purpose: "implement", so the worker keeps its worktree/branch context.
A new title would spawn a fresh worker with no memory of the task. Then
loop to step 1.Closes #N
for its tracker issue and the test-run evidence from step 2: commands
run + results; verify both are present before counting it deliverable).
Record the PR URL in the registry and hand the PR to pr-shepherd.
buhhdy does NOT merge it.claude_code/codex/gemini-*/opencode) dispatched with
purpose: "review" — a DIFFERENT vendor from the one that built the diff. It
reports issues and never edits; only the implementer opens a PR, so a stray
reviewer edit never reaches the deliverable.