agentleFS
lbdremylbdremy/phax/.claude/skills/validation-boundaries/SKILL.md

validation-boundaries

Skill2 starsChanged 5 months agoApache-2.0View on GitHub

Decode every external input (files, env vars, CLI args, API responses) through an Effect Schema before it crosses into the domain.

  • Reads credentials

The file

Files
SKILL.md

Validation boundaries

The rule

Every external input must be decoded through an Effect Schema before crossing into the domain.

External inputs include:

  • Files read from disk (phax.json, phax-plan.json, run-status.json, registry.json, phase/status.json, Claude JSONL output, git command output)
  • Environment variables (process.env.*)
  • CLI arguments (after commander parses them, decode into typed command objects)
  • Any response from an external process

Where decoding happens

Decoding belongs in infra/ adapters or at the very start of cli/ command handlers, before any app/ use case is called. The app/ and domain/ layers receive already-decoded, already-validated values only.

JSON.parse inside app/ or domain/ is a violation. Read the file in infra/, decode immediately, and return a typed value.

Why

Unvalidated data reaching the domain means the domain type system can no longer be trusted. A RunState that was not decoded through its schema might hold an invalid string; a PhaseId built from unvalidated input might be out of range. Effect Schema decoding at the boundary means every value inside the domain is guaranteed valid by construction.

How to fix a violation

  1. Find the JSON.parse (or readFile with an immediate cast) in app/ or domain/.
  2. Move it to the appropriate infra/ adapter.
  3. Decode the raw value through the matching schema in src/schemas/.
  4. Return the typed, decoded value across the boundary.

Audit rule

PHAX_VALIDATION_001 — JSON.parse or unguarded readFile caller detected outside infra/.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.