Exécutez n'importe quel Skill dans Manus
en un clic

Exécutez n'importe quel Skill dans Manus en un clic

yagni-review

Étoiles2

Forks0

Mis à jour23 juin 2026 à 12:29

Review diffs for over-engineering only. Finds what to delete: reinvented standard library, unneeded dependencies, speculative abstractions, dead flexibility. One line per finding with tag and replacement. Use when the user says "review for over-engineering", "what can we delete", "is this over-engineered", "simplify review", "yagni review", or invokes /yagni-review. Complements correctness-focused /review — this one only hunts complexity.

Installation

Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.

Exécuter dans Manus

Source

FranciscoJRSantos

FranciscoJRSantos/personal-agents

Ouvrir le dépôt GitHub Voir les dépôts du créateur

Téléchargement

Exécuter dans Manus

Métiers associésSOC

Basé sur la classification professionnelle SOC

Analystes en assurance qualité des logiciels et testeursProfessions informatiques et mathématiques·SOC 15-1253

SKILL.md

readonly

name

yagni-review

description

YAGNI Review

Review the current branch's diff against main for unnecessary complexity. One line per finding: location, what to cut, what replaces it. The diff's best outcome is getting shorter.

Scope

Over-engineering and complexity only. Correctness bugs, security holes, and performance are explicitly out of scope — route them to /review.

Step 1: Get the Diff

git diff main...HEAD

If main doesn't exist locally, try origin/main, then master.

Step 2: Classify Findings

For each finding, assign one tag:

Tag	What to cut	Replacement
`delete:`	Dead code, unused flexibility, speculative feature	Nothing — remove it
`stdlib:`	Hand-rolled thing the standard library ships	Name the function/class
`native:`	Dependency or code doing what the platform already does	Name the native feature
`yagni:`	Abstraction with one implementation, config nobody sets, layer with one caller	Inline until a second use exists
`shrink:`	Same logic, fewer lines	Show the shorter form

Step 3: Format Output

One line per finding:

L<line>: <tag> <what>. <replacement>.

For multi-file diffs: <file>:L<line>: <tag> <what>. <replacement>.

Examples

L12-38: stdlib: 27-line email validator class. "@" check + confirmation email, 1 line.
L4: native: moment.js imported for one format call. Intl.DateTimeFormat, 0 deps.
repo.py:L88: yagni: AbstractRepository with one implementation. Inline it.
L52-71: delete: retry wrapper around an idempotent local call. Nothing replaces it.
L30-44: shrink: manual loop builds dict. dict(zip(keys, values)), 1 line.

Step 4: Score

End with the only metric that matters:

net: -<N> lines possible.

If there is nothing to cut, say Lean already. Ship. and stop.

Boundaries

Scope: over-engineering and complexity only.
Correctness bugs, security holes, performance are out of scope — route to /review.
A single smoke test or assert-based self-check is the minimum, not bloat — never flag it.
Does not apply the fixes, only lists them.

Optional: Persist Findings

If the user wants to track deferred findings, mark them with deferred: comments in the code (format: # deferred: <ceiling>, <upgrade path>). Then run /debt-ledger to harvest them into a ledger.

Plus depuis ce dépôt

même dépôt

debt-ledger

FranciscoJRSantos/personal-agents

Harvest every `deferred:` comment in the codebase into a debt ledger, so deliberate shortcuts and deferrals get tracked instead of rotting into "later means never". Use when the user says "debt ledger", "/debt-ledger", "what did we defer", "list the shortcuts", "ponytail ledger", or "what did we mark to do later". One-shot report, changes nothing.

2026-06-232

implement

FranciscoJRSantos/personal-agents

Vertical-slice implementation with TDD inside each slice. Reads the plan and kanban-board artifacts to determine which slice to implement. Each slice crosses all layers (data → pure → edge → UI → integration) and follows Red-Green-Refactor. Clear context between slices. Use this skill whenever the user says "implement", "start coding", "proceed with implementation", "/implement", or wants to begin implementation after a plan exists. Entry point: /implement

2026-06-232

grill-me

FranciscoJRSantos/personal-agents

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Produces a decision log artifact consumable by /plan. Invoke with "/grill-me [topic]".

2026-06-072

plan

FranciscoJRSantos/personal-agents

Fetches a Jira ticket, detects its type, and generates a type-aware implementation plan with Kanban/DAG vertical slices. Presents the plan for approval before writing it back to Jira as a comment. Produces both a plan artifact and a kanban-board artifact. Use this skill whenever the user invokes /plan PROJ-123 or asks to generate an implementation plan for a Jira ticket. Entry point: /plan PROJ-123

2026-06-072

ship

FranciscoJRSantos/personal-agents

Run the full verification suite (lint, type check, tests), gate on the review artifact, commit, push, and create a merge request. Use this skill whenever the user says "ship it", "commit and push", "create MR", "open a merge request", "push my branch", or wants to finalize and submit their work.

2026-06-072

learn

FranciscoJRSantos/personal-agents

Codify a correction or convention from the current session into AGENTS.md (global) or .agents/conventions.md (project-specific) in "Always/Never [action] BECAUSE [reason]" format. Use this skill whenever the user says "remember this", "add a rule", "don't do that again", "learn this", "/learn", or wants to capture a lesson from the session before the conversation ends. Entry point: /learn [optional rule text]

2026-05-232

name

yagni-review

description

YAGNI Review

Review the current branch's diff against main for unnecessary complexity. One line per finding: location, what to cut, what replaces it. The diff's best outcome is getting shorter.

Scope

Over-engineering and complexity only. Correctness bugs, security holes, and performance are explicitly out of scope — route them to /review.

Step 1: Get the Diff

git diff main...HEAD

If main doesn't exist locally, try origin/main, then master.

Step 2: Classify Findings

For each finding, assign one tag:

Tag	What to cut	Replacement
`delete:`	Dead code, unused flexibility, speculative feature	Nothing — remove it
`stdlib:`	Hand-rolled thing the standard library ships	Name the function/class
`native:`	Dependency or code doing what the platform already does	Name the native feature
`yagni:`	Abstraction with one implementation, config nobody sets, layer with one caller	Inline until a second use exists
`shrink:`	Same logic, fewer lines	Show the shorter form

Step 3: Format Output

One line per finding:

L<line>: <tag> <what>. <replacement>.

For multi-file diffs: <file>:L<line>: <tag> <what>. <replacement>.

Examples

L12-38: stdlib: 27-line email validator class. "@" check + confirmation email, 1 line.
L4: native: moment.js imported for one format call. Intl.DateTimeFormat, 0 deps.
repo.py:L88: yagni: AbstractRepository with one implementation. Inline it.
L52-71: delete: retry wrapper around an idempotent local call. Nothing replaces it.
L30-44: shrink: manual loop builds dict. dict(zip(keys, values)), 1 line.

Step 4: Score

End with the only metric that matters:

net: -<N> lines possible.

If there is nothing to cut, say Lean already. Ship. and stop.

Boundaries

Scope: over-engineering and complexity only.
Correctness bugs, security holes, performance are out of scope — route to /review.
A single smoke test or assert-based self-check is the minimum, not bloat — never flag it.
Does not apply the fixes, only lists them.

Optional: Persist Findings

If the user wants to track deferred findings, mark them with deferred: comments in the code (format: # deferred: <ceiling>, <upgrade path>). Then run /debt-ledger to harvest them into a ledger.