pos-ai: Command for Claude Code

.claude/commands/eval-loop.md

eval-loop is a command for Claude Code from abuango/pos-ai. It costs 0 tokens per session (643 once invoked), scanned A, original, MIT.

A review loop that checks code changed in the current session, evaluates it against stated criteria, and allows up to three rounds of refinement.

In plain words
What is it for?
Use it after writing or modifying code to inspect the diff, check project rules, and refine issues found during review.
Why use it?
It helps catch quality, security, correctness, and testing problems before the work is considered complete.

Command for Claude Code

Written for Claude Code: installed under .claude/.

This is abuango/pos-ai's own configuration. It tells Claude Code how to work on pos-ai itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything pos-ai configures →

Reuse

Borrowing it

Nothing to install: this file belongs to abuango/pos-ai. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/abuango/pos-ai/main/.claude/commands/eval-loop.md
Clone the repo
git clone --depth 1 https://github.com/abuango/pos-ai

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for eval-loop

README.md
[![agentmods](https://agentmods.dev/badge/commands/abuango/pos-ai/eval-loop/github.svg)](https://agentmods.dev/commands/abuango/pos-ai/eval-loop)
Your own site
<a href="https://agentmods.dev/commands/abuango/pos-ai/eval-loop"><img src="https://agentmods.dev/badge/commands/abuango/pos-ai/eval-loop/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for eval-loop

Your own site · 80×15
<a href="https://agentmods.dev/commands/abuango/pos-ai/eval-loop"><img src="https://agentmods.dev/badge/commands/abuango/pos-ai/eval-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 643 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00000 $0.00643
Opus 5.5 $0.00000 $0.00257
Sonnet 5.5 $0.00000 $0.00129
Haiku 4.5 $0.00000 $0.00064

Measured 2mo ago against content hash f97e9c741b4c, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

eval-loop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/eval-loop.md · 87 lines

How it starts

The opening of the file, as written. The whole thing — 87 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are running an evaluator-optimizer loop on the code that was just written or modified in this session. This implements a structured generate-evaluate-refine cycle with a maximum of 3 iterations.

Process

Step 1: Capture What Changed

Run git diff --unified=3 to see all unstaged changes, or git diff --cached for staged changes. If no diff exists, ask the user which files to evaluate.

Step 2: Evaluate Against Criteria

For each changed file, check these criteria:

Code Quality (from .claude/rules/code-quality.md):

  • No functions longer than 30 lines
  • No logic duplicated more than twice
  • No any type usage in TypeScript
  • Components with 3+ related props use object grouping
  • All async operations have error handling
  • No SELECT * queries
  • Frequently queried columns have indexes

Security (critical — auto-fail):

  • No hardcoded secrets or credentials
  • No SQL injection vectors (raw queries with interpolation)
  • No XSS vectors (innerHTML, v-html, dangerouslySetInnerHTML)
  • No command injection (exec/system with user input)
  • Input validation present at system boundaries

Correctness:

  • Logic matches the approved plan intent
  • Edge cases handled (null, empty, boundary values)
  • No off-by-one errors in loops/pagination
  • Error messages are actionable

Step 3: Report Findings

For each issue found, report:

[FAIL] {criteria} — {file}:{line}
  Issue: {what's wrong}
  Fix: {specific fix}

If all criteria pass:

[PASS] All criteria met. No issues found.

Step 4: Fix and Re-evaluate (if needed)

If issues were found:

  1. Fix each issue using the Edit tool
  2. Re-run the evaluation (Step 2) on the fixed files only
  3. Report iteration results

Iteration tracking:

--- Iteration {n}/3 ---
Issues found: {count}
Issues fixed: {count}
Remaining: {count}

Step 5: Final Report

After all iterations (or when all criteria pass), output:

## Eval-Loop Complete

Iterations: {n}/3
Starting issues: {count}
Resolved: {count}
Remaining: {count}

Status: PASSED / PARTIAL (N issues remain) / FAILED (critical issues)

{List any remaining issues if not fully resolved}

Read the full file on GitHub · 87 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2mo ago First seen · 87 lines · 0 tokens per session scan A f97e9c741b4c

Subscribe to this mod's changes

eval-loop is a command published in the GitHub repository abuango/pos-ai (2 stars, last pushed 7mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 643 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.