findings-diff

Stars4

Forks0

UpdatedJuly 18, 2026 at 15:25

Compare two lean-findings.json files for the same subject across time and report real progress deltas — findings resolved, new, persisting, and severity shifts. Use this skill whenever the user asks "did we improve", "compare this audit to the last one", wants progress since a previous audit, or re-runs an audit on a subject that already has a findings file. Part of Lean Agentic AI Skills; a transformer — emits a diff report plus a merged findings file with status annotations.

Installation

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

Source

navveenb

navveenb/lean-agentic-ai-skills

View GitHub Repository View Creator Repositories

Download

Related occupationsSOC

Based on SOC occupation classification

Software DevelopersComputer and Mathematical Occupations·SOC 15-1252

File Explorer

3 files

SKILL.md

readonly

name

findings-diff

description

Findings Diff

Transformer skill — the suite's progress engine. Input: two findings files for the same subject.identifier (baseline and current). Output: a diff summary + an annotated findings file consumers can render.

This is the one place the module produces legitimate quantitative progress claims: counts of findings resolved, introduced, and persisting between two dated audits are real, countable deltas — no estimation involved.

Procedure

Validate comparability. Same subject identifier and same producer skill(s); if the audits' methodologies differ materially (different scope, different fetch depth), say so up front — a "resolved" finding that merely fell out of scope is not progress. Scope changes go in the diff report's caveats, always.
Match findings by id. IDs are stable slugs by contract. For unmatched-but-similar pairs (same location + same pattern URL, different id), list as "probably renamed" for human confirmation rather than auto-matching.
Classify every finding:
- Resolved — in baseline, not in current (and still in scope).
- New — in current, not in baseline.
- Persisting — in both; note severity or evidence changes.
- Out-of-scope shift — in baseline, but current audit couldn't assess the area. NOT counted as resolved.
Report — a short summary table (resolved/new/persisting by severity), the honest headline ("3 high-severity findings resolved, 1 new medium introduced"), notable evidence changes on persisting items (e.g. image count dropped from 11 to 4 — partial progress is progress), and passed items gained or lost.
Emit the annotated current findings file with a status field per finding so lean-dashboard's trend view and lean-report-writer's progress section can consume it directly.

Honesty rules

Resolved counts are only claimable when scope held constant — the caveat mechanism above is not optional. Never extrapolate deltas into energy/carbon savings; the honest sentence is "N findings resolved; measured impact requires before/after instrumentation." A regression (new highs) is reported as prominently as progress.

Not this skill's job

Producing findings (any producer), rendering the trend UI (lean-dashboard), gating builds on regressions (lean-regression-gate).

More from this repository

same repository

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

findings-diff

description

Findings Diff

Procedure

Validate comparability. Same subject identifier and same producer skill(s); if the audits' methodologies differ materially (different scope, different fetch depth), say so up front — a "resolved" finding that merely fell out of scope is not progress. Scope changes go in the diff report's caveats, always.
Match findings by id. IDs are stable slugs by contract. For unmatched-but-similar pairs (same location + same pattern URL, different id), list as "probably renamed" for human confirmation rather than auto-matching.
Classify every finding:
- Resolved — in baseline, not in current (and still in scope).
- New — in current, not in baseline.
- Persisting — in both; note severity or evidence changes.
- Out-of-scope shift — in baseline, but current audit couldn't assess the area. NOT counted as resolved.
Report — a short summary table (resolved/new/persisting by severity), the honest headline ("3 high-severity findings resolved, 1 new medium introduced"), notable evidence changes on persisting items (e.g. image count dropped from 11 to 4 — partial progress is progress), and passed items gained or lost.
Emit the annotated current findings file with a status field per finding so lean-dashboard's trend view and lean-report-writer's progress section can consume it directly.

Honesty rules

Not this skill's job

Producing findings (any producer), rendering the trend UI (lean-dashboard), gating builds on regressions (lean-regression-gate).