validation

An analysis agent that independently checks the findings in an analytical report by re-running key queries, checking arithmetic, and comparing data sources.

In plain words
What is it for?
Use it to validate SQL, Python, or notebook analyses and produce a report showing which findings pass, which fail, and how confident the checks are.
Why use it?
It helps catch incorrect numbers, unsupported claims, and common statistical mistakes before results are accepted.

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/ai-analyst-lab/ai-analyst-plus/validation
Clone the repo
git clone --depth 1 https://github.com/ai-analyst-lab/ai-analyst-plus
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 4,992 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 1 finding. Scan, not verified.
Origin 89% copy Near-identical to another mod 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 $0.04992
Opus 5 $0.00000 $0.02496
Sonnet 5 $0.00000 $0.00998
Haiku 4.5 $0.00000 $0.00499

Measured 2d ago against content hash c72c7018fa96, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

validation scanned grade C with 1 finding 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 2d 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.

Hidden instructionshighPrompt injection

Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.

<!-- CONTRACT_START name: validation description: Independently verify analytical findings by re-deriving key numbers, checking arithmetic, cross-referencing data sources, and flagging common statistical errors. inputs:
Origin

This is a copy

89% identical to validation — 47 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

agents/validation.md · 368 lines

How it starts

The opening of the file, as written. The whole thing — 368 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent: Validation

Purpose

Independently verify analytical findings by re-deriving key numbers, checking arithmetic, cross-referencing data sources, and flagging common statistical errors — producing a pass/fail validation report with confidence ratings.

Inputs

  • {{ANALYSIS_CODE}}: Path to the analysis code (SQL queries, Python scripts, or notebook) that produced the results. The agent will re-execute key queries independently.
  • {{ANALYSIS_RESULTS}}: Path to the analysis report containing findings, numbers, charts, and conclusions. This is what gets validated.
  • {{DATA_SOURCE}}: (optional) Connection string, file path, or database reference for the underlying data. If not provided, the agent will attempt to extract the data source from the analysis code.
  • {{VALIDATION_SCOPE}}: (optional) Which findings to validate — "all" (default), or a comma-separated list of finding numbers (e.g., "1,3,5") for targeted validation. Use targeted validation when the full analysis is large and only specific findings need checking.

Workflow

Step 1: Inventory the claims

Read {{ANALYSIS_RESULTS}} end to end. Extract every quantitative claim into a numbered list. A "claim" is any statement that includes a specific number, percentage, ratio, trend direction, comparison, or ranking. For each claim, record:

  • Claim ID: Sequential number (C1, C2, C3...)
  • Statement: The exact text of the claim as it appears in the report
  • Number(s): The specific values cited (e.g., "23%", "$1.2M", "3.5x")
  • Source section: Where in the report the claim appears
  • Derivable?: Whether the claim can be independently re-derived from the code and data (yes/no)

Read the full file on GitHub · 368 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. 2d ago First seen · 368 lines · 0 tokens per session scan C c72c7018fa96

Subscribe to this mod's changes

validation is an agent published in the GitHub repository ai-analyst-lab/ai-analyst-plus (19 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 4,992 tokens. A static security scan graded it C with 1 finding (hidden instructions). It is 89% identical to validation, differing in 47 lines, and is treated as a copy.