AGENT SKILL

code-review

Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews in parallel read-only `code-reviewer` sub-agents through the configured review route and reports them side by side. Use when the user wants to review a branch, a PR, work-in-progress changes, or asks to "review since X".

License 许可
MIT
所属仓库
nredd/pi-harness
最近更新
2026-11-15
Stars
★ 2
Forks
1

01 前言 Front Matter

name
code-review
description
Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews in parallel read-only `code-reviewer` sub-agents through the configured review route and reports them side by side. Use when the user wants to review a branch, a PR, work-in-progress changes, or asks to "review since X".

03 SKILL.md 正文

Two-axis review of the diff between HEAD and a fixed point the user supplies:

  • Standards: does the code conform to this repo's documented coding standards?
  • Spec: does the code faithfully implement the originating issue / spec?

Both axes run as parallel code-reviewer sub-agents (read-only, fixed output template; defined by the code-reviewer agent) so they don't pollute each other's context, then this skill aggregates their findings.

The issue tracker should have been provided to you. If docs/agents/issue-tracker.md is missing, tell the user to run /setup-matt-pocock-skills.

Process

1. Pin the fixed point

Whatever the user said is the fixed point (a commit SHA, branch name, tag, main, HEAD~5, etc.). If they didn't specify one, ask for it.

Resolve the fixed point and capture the final full reviewed HEAD SHA before dispatch: git rev-parse <fixed-point> and git rev-parse HEAD. For a non-PR review, use the resolved fixed point as <reviewed-base>. For a PR, run git merge-base <pr-base> HEAD once, record its full SHA as <reviewed-base>, and preserve it across every later HEAD change. Bind the diff command to those immutable values: git diff <reviewed-base>...<reviewed-head> (three-dot, so the comparison is against the original merge-base). Also note the matching commits via git log <reviewed-base>..<reviewed-head> --oneline.

Before going further, confirm the fixed point resolves and the diff is non-empty. A bad ref or empty diff should fail here, not inside two parallel sub-agents. Before aggregating, resolve HEAD again; if it differs from <reviewed-head>, discard both reports and restart from this step with the preserved <reviewed-base>. A corrective-delta review is supplemental only and never replaces either required whole-PR axis.

2. Identify the spec source

Look for the originating spec, in this order:

  1. Issue references in the commit messages (#123, Closes #45, GitLab !67, etc.), fetched via the workflow in docs/agents/issue-tracker.md.
  2. A path the user passed as an argument.
  3. A spec file under docs/, specs/, or .scratch/ matching the branch name or feature.
  4. If nothing is found, ask the user where the spec is. If they confirm there is none, record that verified absence and still invoke the Spec sub-agent.

3. Identify the standards sources

Anything in the repo that documents how code should be written, such as CODING_STANDARDS.md or CONTRIBUTING.md.

On top of whatever the repo documents, the Standards axis always carries the smell baseline below: a fixed set of Fowler code smells (Refactoring, ch.3) that applies even when a repo documents nothing. Two rules bind it:

  • The repo overrides. A documented repo standard always wins; where it endorses something the baseline would flag, suppress the smell.
  • Always a judgement call. Each smell is a labelled heuristic ("possible Feature Envy"), never a hard violation. Like any standard here, skip anything tooling already enforces.

Each smell reads what it is → how to fix; match it against the diff:

  • Mysterious Name: a function, variable, or type whose name doesn't reveal what it does or holds. → rename it; if no honest name comes, the design's murky.
  • Duplicated Code: the same logic shape appears in more than one hunk or file in the change. → extract the shared shape, call it from both.
  • Feature Envy: a method that reaches into another object's data more than its own. → move the method onto the data it envies.
  • Data Clumps: the same few fields or params keep travelling together (a type wanting to be born). → bundle them into one type, pass that.
  • Primitive Obsession: a primitive or string standing in for a domain concept that deserves its own type. → give the concept its own small type.
  • Repeated Switches: the same switch/if-cascade on the same type recurs across the change. → replace with polymorphism, or one map both sites share.
  • Shotgun Surgery: one logical change forces scattered edits across many files in the diff. → gather what changes together into one module.
  • Divergent Change: one file or module is edited for several unrelated reasons. → split so each module changes for one reason.
  • Speculative Generality: abstraction, parameters, or hooks added for needs the spec doesn't have. → delete it; inline back until a real need shows.
  • Message Chains: long a.b().c().d() navigation the caller shouldn't depend on. → hide the walk behind one method on the first object.
  • Middle Man: a class or function that mostly just delegates onward. → cut it, call the real target direct.
  • Refused Bequest: a subclass or implementer that ignores or overrides most of what it inherits. → drop the inheritance, use composition.

4. Spawn both sub-agents in parallel

A top-level parent spawns two Agent calls in one message, both subagent_type: "code-reviewer", run_in_background: true, and no model override. A delegated session omits run_in_background, uses the supported blocking calls, and consumes both inline results before returning. The configured agentModels route selects the reviewer model. The agent is read-only and already carries the output template, the hard/judgement rule and the word cap: do NOT restate them. Each prompt names its axis and supplies only the axis material. If the review route is quota-parked, wait; do not silently substitute a model.

Standards prompt (Axis: Standards) includes:

  • The full original reviewed-base SHA, full reviewed HEAD SHA, and immutable whole-PR diff command and commit list bound to both.
  • The list of standards-source files you found in step 3, plus the smell baseline from step 3 pasted in full (the sub-agent has no other access to it).

Spec prompt (Axis: Spec) includes:

  • The full original reviewed-base SHA, full reviewed HEAD SHA, and immutable whole-PR diff command and commit list bound to both.
  • The path or fetched contents of the spec, with line numbers so findings can quote them.

Always invoke the Spec sub-agent. When absence is verified in step 2, explicitly supply "No spec available." and the full reviewed HEAD so it returns the authentic SHA-bound no-spec skeleton. An unlocated spec without confirmation is not verified absence; ask before dispatch. A delegated worker asks its parent for that confirmation rather than claiming absence.

If a sub-agent returns something that does not match the template (including a bare verdict with no Checked: list), report it verbatim under its axis, say so, and re-run that axis; do not reformat or re-derive findings.

5. Aggregate

Present the two reports under ## Standards and ## Spec headings, verbatim or lightly cleaned. Do not merge or rerank findings, because the two axes are deliberately separate (see Why two axes).

End with a one-line summary: total findings per axis, and the worst issue within each axis (if any). Don't pick a single winner across axes: that's the reranking the separation exists to prevent.

Why two axes

A change can pass one axis and fail the other:

  • Code that follows every standard but implements the wrong thing → Standards pass, Spec fail.
  • Code that does exactly what the issue asked but breaks the project's conventions → Spec pass, Standards fail.

Reporting them separately stops one axis from masking the other.

04 同仓库更多 Skills