codeclone-hotspots

스타12

포크2

업데이트2026년 6월 17일 18:22

Fast CodeClone quality snapshot — health, top risks, or one metric, without a full review loop.

설치

Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.

출처

orenlab

orenlab/codeclone

GitHub 저장소 열기 Creator 저장소 보기

다운로드

CodeClone Hotspots

Fast quality answer (health, worst hotspots, one metric) — not a full review session.

When to use

"How healthy is this repo?" / "Worst hotspots?" / "Complexity hotspots?" / pre-merge sanity.
Baseline-relative new/known is NOT patch-local proof — for "did my patch cause this?" use the change-control before→after verify path.

Loop

analyze_repository(root=<abs>) → get_production_triage

Cheapest useful path. Stop there unless asked for more.

Specific metric: analyze_repository → check_complexity | check_coupling | check_cohesion | check_dead_code | check_clones
Adoption / API surface / coverage join: get_report_section(section="metrics") (coverage unclear → help(topic="coverage"))
Gate preview → its findings: evaluate_gates(run_id, fail_on_new=, fail_complexity=, …) → for the actual findings, list_findings(novelty="new", family=…, source_kind="production").

Reading the response

Key / easily-misread fields; the real response carries more.

Field	Meaning
`health.score`/`grade`	0–100 / A–F; `dimensions` = per-family
`findings.new`/`known`	baseline-relative — NOT patch-local proof
`new_by_source_kind`	new split prod / tests / fixtures (the gate counts production)
`evaluate_gates.would_fail` + `reasons[]`	gate verdict + cause tokens
check item `novelty` / `clone_type` / `scope`	new vs known / Type-1..4 / production vs tests

Rules

MCP tools only (CodeClone plugin). Absolute root. No latest run → analyze_repository first.
Default thresholds — this is a quick check. detail_level for lists: summary | normal | full.
One precise call beats three. Summarize concisely — a snapshot, not a report.
Do not fall back to CLI / local report files. CodeClone is the source of truth.
If results look concerning, suggest codeclone-review for a real session.

이 저장소의 다른 Skills

같은 저장소

codeclone-setup

orenlab/codeclone

Human repository setup readiness — status, plan, apply, and wizard over the CLI (not MCP).

2026-07-0712

codeclone-architecture-triage

orenlab/codeclone

Rank demonstrated architectural problems from one canonical run plus targeted source verification. Read-only; priorities are response-local heuristics, never CodeClone findings.

2026-06-2712

codeclone-blast-radius

orenlab/codeclone

Inspect structural blast radius before editing — dependents, clone cohort, risk signals, do-not-touch boundaries. Read-only; does not declare intent.

2026-06-2712

codeclone-change-control

orenlab/codeclone

MANDATORY HARD GATE when CodeClone MCP is connected. Before ANY repository file write, obtain edit_allowed=true from start_controlled_change. Before claiming done/verified/ready, finish_controlled_change must accept and clear the intent.

2026-06-2712

codeclone-engineering-memory

orenlab/codeclone

Use CodeClone Engineering Memory through MCP to retrieve scoped evidence before edits and preserve durable incidents, decisions, risks, contradictions, and verification anchors as draft candidates. Memory never grants edit permission; agents propose, humans govern.

2026-06-2712

codeclone-implementation-context

orenlab/codeclone

Before broad rg/grep, non-trivial planning, unclear scope, contract/schema work, or resuming another agent — call get_implementation_context once to bound the implementation frontier. Then use targeted source reads. Read-only; never grants edit permission.

2026-06-2712

name	codeclone-hotspots
description	Fast CodeClone quality snapshot — health, top risks, or one metric, without a full review loop.

CodeClone Hotspots

Fast quality answer (health, worst hotspots, one metric) — not a full review session.

When to use

"How healthy is this repo?" / "Worst hotspots?" / "Complexity hotspots?" / pre-merge sanity.
Baseline-relative new/known is NOT patch-local proof — for "did my patch cause this?" use the change-control before→after verify path.

Loop

analyze_repository(root=<abs>) → get_production_triage

Cheapest useful path. Stop there unless asked for more.

Specific metric: analyze_repository → check_complexity | check_coupling | check_cohesion | check_dead_code | check_clones
Adoption / API surface / coverage join: get_report_section(section="metrics") (coverage unclear → help(topic="coverage"))
Gate preview → its findings: evaluate_gates(run_id, fail_on_new=, fail_complexity=, …) → for the actual findings, list_findings(novelty="new", family=…, source_kind="production").

Reading the response

Key / easily-misread fields; the real response carries more.

Field	Meaning
`health.score`/`grade`	0–100 / A–F; `dimensions` = per-family
`findings.new`/`known`	baseline-relative — NOT patch-local proof
`new_by_source_kind`	new split prod / tests / fixtures (the gate counts production)
`evaluate_gates.would_fail` + `reasons[]`	gate verdict + cause tokens
check item `novelty` / `clone_type` / `scope`	new vs known / Type-1..4 / production vs tests

Rules

MCP tools only (CodeClone plugin). Absolute root. No latest run → analyze_repository first.
Default thresholds — this is a quick check. detail_level for lists: summary | normal | full.
One precise call beats three. Summarize concisely — a snapshot, not a report.
Do not fall back to CLI / local report files. CodeClone is the source of truth.
If results look concerning, suggest codeclone-review for a real session.