lean-regression-gate

스타4

포크0

업데이트2026년 7월 18일 15:25

Gate CI on sustainability regressions — compare a fresh audit's findings against a committed baseline findings file and fail the build (or warn) when new high-severity findings appear. Use this skill whenever the user wants sustainability checks in CI/CD, asks to "fail the build if we get less green", wants a merge check for efficiency regressions, or mentions enforcing green budgets automatically. Part of Lean Agentic AI Skills; a transformer designed to run headless in pipelines.

설치

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

출처

navveenb

navveenb/lean-agentic-ai-skills

GitHub 저장소 열기 Creator 저장소 보기

다운로드

Green Regression Gate

Transformer skill built for CI. Input: a baseline lean-findings.json (committed to the repo) + a fresh audit of the same subject. Output: a pass/warn/fail verdict with a machine-readable summary and a human-readable comment block (for PR comments).

Performance budgets made speed a merge-time concern; this makes efficiency one. The gate's job is narrow: detect regressions against an agreed baseline — it is not a linter that blocks on pre-existing debt.

Gate policy (default, overridable per repo)

FAIL — any new high-severity finding (present in fresh audit, absent from baseline).
WARN — new medium findings; severity escalations on persisting findings; loss of a passed item.
PASS — no new findings at/above threshold; resolved findings noted with congratulations kept to one line.
NEUTRAL/SKIP — scope mismatch between baseline and fresh audit (different methodology depth): the gate must skip with an explanation rather than fail on incomparable data. A gate that fails for tooling reasons erodes trust in one sprint.

Procedure

Reuse findings-diff mechanics for matching and classification (same id-matching and scope-comparability rules).
Apply the policy. Emit: exit-code semantics (0 pass, 1 fail, 78 skip), a compact JSON verdict ({verdict, new_findings[], escalations[], resolved[]}), and a markdown comment block listing each blocking finding with its evidence and fix — actionable in the PR, not just red.
Baseline lifecycle: on intentional acceptance of new debt, the fix is updating the committed baseline in the same PR with a stated reason — the gate should explain this path in its failure message so teams aren't stuck.

Setup guidance (when asked to design the pipeline step)

Producer audit runs first in CI (e.g. web-carbon-audit on the built site preview, lean-ci-audit on workflow changes), gate runs second; baseline lives at a conventional path (.green/baseline.lean-findings.json). Keep the audited subject deterministic (built artifacts, not live prod) so the gate is reproducible.

Honesty rules

The gate never blocks on directional-impact wording or unmeasured claims — only on the presence/severity of findings, which are evidence-backed by contract. False-positive escapes matter: the skip-on-scope-mismatch rule and the baseline-update path are mandatory parts of any setup this skill designs.

Not this skill's job

Producing the audit (producers), longitudinal reporting (findings-diff + lean-dashboard), CI pipeline efficiency itself (lean-ci-audit).

이 저장소의 다른 Skills

같은 저장소

agentic-loop-auditor

navveenb/lean-agentic-ai-skills

Audit agent frameworks and agentic workflows for token-burning loop pathologies — unbounded retries, excessive planning depth, tool-call thrashing, context snowballing, missing loop budgets and kill switches. Use this skill whenever the user shares agent configs or traces (LangChain/LangGraph, CrewAI, AutoGen, custom loops), complains an agent is slow or expensive, or is designing any autonomous/multi-step agent. Part of Lean Agentic AI Skills; emits lean-findings.json.

2026-07-184

ai-eval-efficiency-audit

navveenb/lean-agentic-ai-skills

Audit AI evaluation infrastructure for wasted compute — full benchmark suites re-run on unchanged cases, LLM-as-judge grading without caching, oversized judge models, redundant eval passes per commit, and missing result reuse. Use this skill whenever the user shares eval harness configs or CI eval steps, complains that evals are slow or expensive, mentions LLM-as-judge costs, or runs benchmark suites on every change. Part of Lean Agentic AI Skills; emits lean-findings.json.

2026-07-184

api-efficiency-audit

navveenb/lean-agentic-ai-skills

Audit API design for chattiness and over-fetching — REST endpoints requiring N calls per screen, GraphQL resolvers with N+1 fan-out, default-everything payloads, missing pagination and field selection, provider-side polling where webhooks serve, and uncompressed responses. Use this skill whenever the user shares API schemas (OpenAPI, GraphQL SDL), resolver code, endpoint designs, or asks why clients make too many calls or payloads are heavy. Part of Lean Agentic AI Skills; emits lean-findings.json.

2026-07-184

caching-strategy-designer

navveenb/lean-agentic-ai-skills

Design caching layers so work is done once, not per request — HTTP/CDN caching, application-level caches (Redis/Memcached), computed-result caching, and cache-invalidation strategy. Use this skill whenever the user wants to add caching, asks why the same work is recomputed, designs a read-heavy system, or an audit skill (db-efficiency-audit, web-carbon-audit) flags uncached repeated work. Part of Lean Agentic AI Skills; emits a cache design plus lean-findings.json.

2026-07-184

carbon-aware-scheduler

navveenb/lean-agentic-ai-skills

Design time-shifting of flexible workloads (batch jobs, CI, training, backups, crons) to hours and regions when grid electricity is cleanest, using carbon-intensity signals like Electricity Maps, WattTime, or the Carbon Aware SDK. Use this skill whenever the user mentions carbon-aware computing, scheduling flexible/deferrable jobs, demand shifting, running workloads when energy is clean, or asks how to reduce the carbon intensity (not just amount) of their compute. Part of Lean Agentic AI Skills; emits schedule designs plus lean-findings.json.

2026-07-184

carbon-budget-designer

navveenb/lean-agentic-ai-skills

Design per-feature and per-service efficiency budgets — page-weight budgets, token budgets per AI feature, query budgets per endpoint, GPU-hour budgets per training project — with owners, review checkpoints, and enforcement hooks. Use this skill whenever the user wants sustainability targets for teams, asks how to keep efficiency from regressing organizationally, mentions performance-budget-style governance for carbon/energy, or is setting engineering standards. Part of Lean Agentic AI Skills; an advisor emitting budget definitions plus lean-findings.json.

2026-07-184

name

lean-regression-gate

description

Green Regression Gate

Gate policy (default, overridable per repo)

FAIL — any new high-severity finding (present in fresh audit, absent from baseline).
WARN — new medium findings; severity escalations on persisting findings; loss of a passed item.
PASS — no new findings at/above threshold; resolved findings noted with congratulations kept to one line.
NEUTRAL/SKIP — scope mismatch between baseline and fresh audit (different methodology depth): the gate must skip with an explanation rather than fail on incomparable data. A gate that fails for tooling reasons erodes trust in one sprint.

Procedure

Reuse findings-diff mechanics for matching and classification (same id-matching and scope-comparability rules).
Apply the policy. Emit: exit-code semantics (0 pass, 1 fail, 78 skip), a compact JSON verdict ({verdict, new_findings[], escalations[], resolved[]}), and a markdown comment block listing each blocking finding with its evidence and fix — actionable in the PR, not just red.
Baseline lifecycle: on intentional acceptance of new debt, the fix is updating the committed baseline in the same PR with a stated reason — the gate should explain this path in its failure message so teams aren't stuck.

Setup guidance (when asked to design the pipeline step)

Honesty rules

Not this skill's job

Producing the audit (producers), longitudinal reporting (findings-diff + lean-dashboard), CI pipeline efficiency itself (lean-ci-audit).