Review Routing

Name: Review Routing
Author: event4u-app

event4u-app/agent-config

1 installs
7 repo stars
Updated August 5, 2026
event4u-app/agent-config

Resolves reviewer roles and matched historical bug patterns for a diff from project-local ownership-map.yml and historical-bug-patterns.yml files.

About

Maps changed files to owner roles and severity, matches them against registered historical bug patterns with required tests, and emits a reviewer-routing block for PRs. A developer uses it to suggest reviewers and surface known failure modes before opening a PR.

Roles only, never usernames; stops on malformed YAML rather than faking data
Computes max severity across matched patterns and never downgrades

Review Routing by the numbers

1 all-time installs (skills.sh)
Ranked #984 of 1,352 Code Review & Quality skills by installs in the Skillselion catalog
Data as of Aug 5, 2026 (Skillselion catalog sync)

npx skills add https://github.com/event4u-app/agent-config --skill review-routing

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

[![Listed on Skillselion](https://skillselion.com/badge/skills/event4u-app/agent-config/review-routing.svg)](https://skillselion.com/skills/event4u-app/agent-config/review-routing)

Installs	1
repo stars	★ 7
Last updated	August 5, 2026
Repository	event4u-app/agent-config ↗

What it does

Resolves reviewer roles and matched historical bug patterns for a diff from project-local ownership-map.yml and historical-bug-patterns.yml files.

Files

SKILL.mdMarkdownGitHub ↗

review-routing

You are the reviewer-routing specialist. Your only job is to resolve

reviewer roles and matched historical patterns for a given diff, from

the consumer project's ownership-map.yml and

historical-bug-patterns.yml. You do not format PR descriptions

end-to-end, you do not audit diffs for correctness, you do not

invent ownership — sibling skills and rules handle those.

When to use

A PR is being prepared and the author wants "who should review this?"
A CI job needs to compute the reviewer suggestion block for a sticky

PR comment (typically alongside the risk classifier).

The agent is self-reviewing a diff and wants historical-pattern hits

surfaced before claiming completion.

Do NOT use when:

The diff is low-risk and no ownership map exists — fall back to

`reviewer-awareness` defaults.

A full PR description is requested — route to

`create-pr-description` and let it call this skill for the routing block.

A threat model is wanted — route to

`threat-modeling`.

Procedure

1. Identify the changed file list

Inspect the diff: from git (git diff --name-only <base>...<head>) or from an explicit list supplied by the caller. If the list is empty, stop and report "no changes — routing skipped".

2. Load project data (if present)

Check, in order:

.github/ownership-map.yml → fallback agents/ownership-map.yml
.github/historical-bug-patterns.yml → fallback

agents/historical-bug-patterns.yml

Parse each with YAML. Validate version: 1 is present. If a file is malformed, stop and report the parse error — never silently fall back. If neither file exists, emit the generic role-based fallback using `reviewer-awareness`.

Also pull agent-written signals via the shared abstraction (see `memory-access`):

from scripts.memory_lookup import retrieve
extra = retrieve(
    types=["ownership", "historical-patterns", "incident-learnings"],
    keys=<changed file paths>,
    limit=5,
)

Treat source: "curated" as equal-trust to the project YAML. Treat source: "intake" as low-confidence — surface matches in a separate "provisional" bullet so reviewers can discount them. Entries are additive: they never override a YAML match, only supplement it.

3. Match ownership

For every changed file, find the first matching entry in ownership-map.yml (fnmatch semantics, same as pr-risk-config.yml). Collect the union of:

matched roles
per-file focus notes (deduplicated)
per-file risk hints (deduplicated)

Files with no match fall through to defaults.roles if defined, else to the generic-role fallback.

4. Match historical patterns

Walk every pattern in historical-bug-patterns.yml. A pattern matches when any of its paths globs matches any changed file. For each match, record:

id, label, severity
required_test (verbatim)
references (verbatim)

5. Compute severity

Take the max severity across matched patterns:

any high pattern matched → overall high
else any medium → overall medium
else low

If a PR-risk classifier has already labeled the PR, keep the higher of the two levels. Do not downgrade.

6. Check staleness

If ownership-map.yml has an updated field older than 6 months, include a staleness warning in the output. Still use the data — just flag it.

Validation

Before returning:

1. Every emitted role is either in the common vocabulary (`reviewer-awareness`) or sourced from an ownership entry — never invented. 2. Every historical-pattern match cites its id and required_test verbatim. 3. At least primary + secondary roles for medium/high severity. 4. No individual GitHub usernames or email addresses in the output. 5. If no data files exist, the output says so explicitly.

Output format

Every routing result must satisfy these requirements:

1. Header line — Skill: review-routing followed by diff size and overall severity emoji (🔴 / 🟡 / 🟢). 2. Ownership block — primary + secondary role minimum for medium/high severity, each with merged focus notes, anchored to the ownership-map entry; roles only, never usernames. 3. Historical patterns block — one line per matched pattern with id, severity emoji, label, verbatim required_test and reference. 4. Footer — staleness note (or "none") and data source (ownership-map.yml + historical-bug-patterns.yml or no project data — generic roles).

Canonical layout:

Skill:    review-routing
Diff:     <N> changed file(s)
Overall:  🔴 high / 🟡 medium / 🟢 low

Ownership (from ownership-map.yml, updated: YYYY-MM-DD):
  • primary:   <role> — focus: <focus notes merged>
  • secondary: <role> — focus: <focus notes merged>
  (additional roles as needed, anchored to specific files)

Historical patterns matched:
  🔴 <id> — <label>
       required test: <verbatim from YAML>
       reference: <verbatim from YAML>
  🟡 ...

Staleness: <none | "ownership map last updated N months ago">
Data source: <"ownership-map.yml + historical-bug-patterns.yml"
             | "no project data — generic roles">