# Repo Review

> Use when a local worktree/index or immutable snapshot, range, pull request, release candidate, or verified package needs read-only review; own local commit-readiness and exact staging guidance or immutable severity-ranked findings.

- **Type:** Skill
- **Install:** `agentstack add skill-idaibin-skills-repo-review`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [idaibin](https://agentstack.voostack.com/s/idaibin)
- **Installs:** 0
- **Category:** [AI & ML](https://agentstack.voostack.com/c/ai-and-ml)
- **Latest version:** 0.1.0
- **License:** Apache-2.0
- **Upstream author:** [idaibin](https://github.com/idaibin)
- **Source:** https://github.com/idaibin/skills/tree/main/skills/repo-review
- **Website:** https://skills.sh/idaibin/skills

## Install

```sh
agentstack add skill-idaibin-skills-repo-review
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Repository Review

## Overview

Review repository changes without modifying files, Git, GitHub, or remote state. Select the review basis first, then apply the matching mode. Worktree mode owns dirty-tree classification and commit-readiness guidance; immutable modes own fixed-SHA P0-P3 review. All modes share contract tracing, structural completeness, specialist coordination, evidence quality, and read-only authority.

## Review Basis

Select exactly one basis before conclusions:

- **Worktree/index:** current tracked, untracked, staged, and unstaged state.
- **Repository snapshot:** one tag, branch resolved to SHA, or commit.
- **Commit range:** explicit immutable `base..head` SHAs.
- **Pull request:** metadata plus complete changed-file evidence and resolved base/head SHAs.
- **Release candidate:** immutable candidate revision plus release/config/CI evidence.
- **Review package:** verified manifest, hashes, coverage, exclusions, and final marker.

Do not mix evidence between bases. Current-worktree content is contamination when reviewing another SHA unless explicitly included in the basis.

## Workflow

1. Read effective repository guidance and record the requested review object, scope, output, and non-goals.
2. Fix the basis before conclusions:
   - for Worktree, run status, diff stat/name-status, and cached equivalents when staged content exists;
   - for immutable modes, resolve full SHAs or verify the package manifest and hashes.
3. Build the smallest complete read set from changed or explicitly owned paths. A `repo-map` artifact may guide navigation but is never review proof.
4. In Worktree mode, classify every changed file as `task-owned`, `related-existing`, `unrelated-existing`, `mixed-hunk`, or `unknown` before proposing commit groups.
5. Trace relevant interfaces through registrations, callers, types, data shaping, persistence, generated artifacts, runtime config, tests, docs, CI/deploy, and stale references.
6. Evaluate two independent axes:
   - **Standards:** repository guidance, architecture, correctness, security, performance, maintainability, and applicable domain conventions.
   - **Spec:** originating requirements, decisions, acceptance criteria, missing behavior, wrong behavior, and unrequested scope.
   If no trustworthy spec exists, mark Spec `Not verified`; do not infer one from the diff.
7. Keep the two evidence passes independent. They may run in parallel only when delegation is available, both scopes are read-only and fixed, and the coordinator can verify and integrate their results.
8. Select only applicable frontend, Rust, security, validation, and release profiles. Delegate bounded evidence to `audit-frontend`, `audit-rust`, or `audit-security`; retain scope, integration, deduplication, severity, and final report ownership.
9. Resolve documented path mismatches at the selected basis. If a path or parent is absent, ascend to the nearest existing ancestor and search only the relevant subtree; route repo-map edits to `repo-map`.
10. Reject speculative, unreachable, style-only, duplicate, or already-resolved findings. Consolidate both axes into P0-P3 findings from concrete impact and urgency while retaining each finding's axis.
11. Run only non-mutating repository checks needed for the selected basis and risk.
12. In Worktree mode, produce semantic commit groups and exact path- or hunk-level staging guidance for `repo-delivery`. In immutable modes, report merge/release implications without staging guidance.
13. Report exclusions, residual risks, failed checks, and every `Not found` or `Not verified` gap.

## Modes

- **Worktree review:** dirty-tree ownership, mixed hunks, contract completeness, commit readiness, semantic groups, exact staging guidance, and commit messages.
- **Snapshot review:** bounded review of one fixed repository revision.
- **Range review:** `base..head` regressions, contracts, tests, docs, and structural lifecycle.
- **Pull-request review:** complete PR evidence without posting comments unless separately authorized.
- **Release review:** compatibility, migrations, generated artifacts, packaging, CI, deployment, rollback, and security-sensitive configuration.
- **Review-package assessment:** package integrity and evidence coverage before findings.

## Do Not Use For

- Repository mapping or repo-map maintenance; use `repo-map`.
- Future implementation planning; use the host's built-in planning.
- Business-domain modeling without a change basis; use `domain-modeling`.
- Root-cause diagnosis of a concrete failure; use the host's built-in diagnosis under effective instructions.
- A bounded security-only audit when the target surface is already known; use `audit-security`.
- Implementing accepted fixes; use the matching `implement-*` skill.
- Staging, commits, pushes, squash, cleanup, or other Git mutation; use `repo-delivery` after explicit authorization.
- External ChatGPT sending or browser/client operation; use the matching operations skill.

## Hard Rules

- Keep every review read-only. Do not edit, format, stage, unstage, commit, push, change refs, post comments, or create issues/PRs.
- State the review basis and resolved SHAs before immutable conclusions; state complete status/index evidence before Worktree conclusions.
- Never use current-worktree files to clear a finding at another SHA.
- Inventory the full dirty tree even when Worktree scope is narrower, then preserve unrelated changes.
- Mark mixed files `mixed-hunk`; never recommend whole-file staging unless every hunk belongs to the group.
- Do not recommend `git add .`, `git add -A`, directory-wide adds, or broad wildcards unless explicitly approved.
- Do not claim whole-repository, PR, release, or package coverage from partial evidence.
- Do not report findings without reachable evidence and concrete impact.
- Do not approve structural add/reuse/move/rename/delete work while manifests, exports, commands, tests, CI/deploy, docs, indexes, migrations, generated files, consumers, or stale references disagree.
- Treat runtime, CI, deployment, external services, branch policy, and package completeness as `Not verified` unless directly evidenced.

## Output Contract

Lead with mode, basis, scope, exclusions, and validation, then severity-ranked P0-P3 findings labeled `Standards`, `Spec`, or both. Every finding includes exact location, cited standard or requirement when available, evidence, impact, remediation, and verification. Include a concise Standards verdict and Spec Compliance verdict; mark missing specification evidence `Not verified`. For Worktree mode, also report ownership labels, staged-state risks, semantic commit groups, exact staging approach, validation status, and concise commit messages. For immutable modes, include resolved SHAs, cross-domain integration, merge/release implications, and `No actionable findings` for empty requested groups. Finish with final verdict, residual risk, and all `Not found` or `Not verified` items.

## References

- See [references/usage.md](references/usage.md) for routing and mode examples.
- See [references/worktree-checklist.md](references/worktree-checklist.md) for dirty-tree ownership and commit-readiness review.
- See [references/checklist.md](references/checklist.md) for immutable basis, severity, and release review.
- See [references/standards-and-spec.md](references/standards-and-spec.md) for independent Standards and Spec review axes.
- See [references/worktree-examples.md](references/worktree-examples.md) for commit grouping examples.
- See [references/eval-cases.md](references/eval-cases.md) for trigger, boundary, scenario, and quality evals.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [idaibin](https://github.com/idaibin)
- **Source:** [idaibin/skills](https://github.com/idaibin/skills)
- **License:** Apache-2.0
- **Homepage:** https://skills.sh/idaibin/skills

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-idaibin-skills-repo-review
- Seller: https://agentstack.voostack.com/s/idaibin
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.
