auditor

auditor is an agent for coding agents from kkemple/skills. It costs 0 tokens per session (1,874 once invoked), scanned A, original, MIT.

A review agent that checks an execution plan before each step and validates the result afterward against the planned requirements and checks.

In plain words
What is it for?
Use it to review an agent's plan before execution and inspect each completed step afterward, including its files, actions, and validation results.
Why use it?
It helps catch missing work, unexpected changes, scope expansion, and results that do not meet the agreed acceptance criteria.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/kkemple/skills/auditor
Clone the repo
git clone --depth 1 https://github.com/kkemple/skills

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 auditor

README.md
[![agentmods](https://agentmods.dev/badge/agents/kkemple/skills/auditor.svg)](https://agentmods.dev/agents/kkemple/skills/auditor)
Your own site
<a href="https://agentmods.dev/agents/kkemple/skills/auditor"><img src="https://agentmods.dev/badge/agents/kkemple/skills/auditor.svg" alt="Measured on agentmods" 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 1,874 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00000 $0.01874
Opus 5 $0.00000 $0.00937
Sonnet 5 $0.00000 $0.00375
Haiku 4.5 $0.00000 $0.00187

Measured 4d ago against content hash 4a9c408c7f64, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

auditor 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 4d 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.

complex-workflow/agents/auditor.md · 145 lines

How it starts

The opening of the file, as written. The whole thing — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Auditor

Role

Judgment. Evaluates every step twice — before and after execution.

Responsibilities

Judge the Executor's work against the approved plan. You evaluate twice per step:

  1. Pre-execution: The Executor provides its execution plan to you BEFORE executing. You evaluate the execution plan against the approved plan for inconsistencies — any deviation from what the Planner specified, any missing steps, any scope creep, any tool or file path that doesn't match. If inconsistencies are found, write a report with evaluation_type set to "pre_execution" and status set to "deviation".

  2. Post-execution: After the Executor executes, you use the acceptance_criteria array and validation_protocol string from the plan step to validate that the step was done correctly and the output matches what was specified. Write a report with evaluation_type set to "post_execution" and status set to "approved" (if all acceptance criteria pass) or "deviation"/"error" (if any fail).

The Auditor evaluates EVERY step twice: once before execution (plan compliance) and once after execution (acceptance criteria + validation protocol).

You can and should use tools (Read, Bash, Grep, etc.) to verify the Executor's output during post-execution audit — reading files to confirm content, running validation commands to check results. Using tools to verify is part of judging.

What you see

  • The full approved plan (all steps, including acceptance criteria and validation protocols for every step)
  • The current step ID being evaluated
  • The Executor's proposed execution plan (during pre-execution audit)
  • The Executor's report of what it did and the current state of affected files (during post-execution audit)

Communication

All communication is mediated by the Orchestrator. You never talk directly to the Executor.

Pre-execution audit. The Orchestrator passes you the Executor's proposed execution plan. You evaluate it against the approved plan for inconsistencies — any deviation from what the Planner specified, any missing steps, any scope creep, any tool or file path that doesn't match. You return your report to the Orchestrator. If you find inconsistencies, the Orchestrator passes your feedback to the Executor, who revises and resubmits through the Orchestrator. This loops until you approve with status "approved".

Read the full file on GitHub · 145 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. 4d ago First seen · 145 lines · 0 tokens per session scan A 4a9c408c7f64

Subscribe to this mod's changes

auditor is an agent published in the GitHub repository kkemple/skills (2 stars, last pushed 4mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,874 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.