environment-config-auditor

SkillFiles & storage

Reviews .env/config files for consistency and missing required variables across environments. Use when comparing dev/staging/prod config, or debugging an "it works locally but not in prod" issue.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the environment-config-auditor skill

What this skill tells your AI

The instructions your AI receives, as published by codebygarv/ai-skills in skills/utilities/environment-config-auditor/SKILL.md and read by ahel’s review.

Purpose

Review environment/config files (.env, YAML/JSON config, etc.) for consistency across environments — missing variables, inconsistent naming, and unexplained differences that commonly cause "works locally, breaks in prod" bugs.

When to Use

  • Comparing config across environments (dev/staging/prod) to spot drift.
  • Debugging an issue that only happens in one environment.
  • Onboarding a new environment and want to confirm nothing's missing relative to existing ones.

What to Analyze

  1. Missing variables — a variable present in one environment's config but absent in another, where that's not obviously intentional (e.g. a DEBUG flag being dev-only is fine; a DATABASE_URL missing in prod is not).
  2. Naming inconsistency — the same concept named differently across files (DB_URL in one, DATABASE_URL in another) — a common source of "I set the variable but it's not being picked up" bugs.
  3. Type/format inconsistency — a boolean represented as "true" in one environment and 1 in another, where the parsing code might not handle both.
  4. Hardcoded values that should be per-environment — the same value (e.g. an API URL) present identically across all environments, when it plausibly should differ (pointing to a prod API from a dev config).
  5. Values that look like leftover secrets — flag (don't just silently note) anything that looks like it could be a real credential rather than a placeholder — cross-reference with Secrets Scanner territory if found.
  6. Unused variables — defined in config but not obviously referenced anywhere the reviewer can see (flag as "worth checking," since absence of visible usage isn't proof of non-use without full codebase access).

Output Format

  • Findings grouped by category (Missing Variables, Naming Inconsistency, Type/Format Inconsistency, Suspicious Hardcoded Values, Possible Secrets, Possibly Unused).
  • Each finding: the variable(s) involved, which environment(s), and the concrete concern.
  • Flag anything that looks like a real secret as top priority, with a pointer to Secrets Scanner for a dedicated deeper pass.

Avoid

  • Flagging intentional environment-specific differences (like debug flags, feature flags) as if they were bugs — distinguish "different on purpose" from "different by accident" where inferable from naming/context.
  • Assuming a variable is unused without caveat — this skill sees only the config files given, not the full codebase's usage.
  • Repeating a found secret value in plaintext in the output — redact, same as Secrets Scanner.

Signals

GitHub stars
25
Forks
1
Last commit
Aug 2026
Advanced
Catalog kind
skill
Gateway key
environment-config-auditor
Source
github.com/codebygarv/ai-skills