orchestrate-multimodal-auditor

orchestrate-multimodal-auditor is a skill for Claude Code, Codex from NITISH-R-G/hackerrank-orchestrate-skills. It costs 55 tokens per session (701 once invoked), scanned A, original, MIT.

A test for systems that use more than one kind of input, such as text plus images or audio, to make decisions.

In plain words
What is it for?
Use it to compare decisions with a media input enabled and disabled, and document whether the modality changes routing.
Why use it?
It shows whether the image or audio actually affects the result rather than merely being processed in the background.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it to compare decisions with a media input enabled and disabled, and document whether the modality changes routing.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/nitish-r-g/hackerrank-orchestrate-skills/orchestrate-multimodal-auditor
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.

Any agent
npx skills add NITISH-R-G/hackerrank-orchestrate-skills --skill orchestrate-multimodal-auditor
Clone the repo
git clone --depth 1 https://github.com/NITISH-R-G/hackerrank-orchestrate-skills

Made for: Claude Code, Codex.

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 orchestrate-multimodal-auditor

README.md
[![agentmods](https://agentmods.dev/badge/skills/nitish-r-g/hackerrank-orchestrate-skills/orchestrate-multimodal-auditor/github.svg)](https://agentmods.dev/skills/nitish-r-g/hackerrank-orchestrate-skills/orchestrate-multimodal-auditor)
Your own site
<a href="https://agentmods.dev/skills/nitish-r-g/hackerrank-orchestrate-skills/orchestrate-multimodal-auditor"><img src="https://agentmods.dev/badge/skills/nitish-r-g/hackerrank-orchestrate-skills/orchestrate-multimodal-auditor/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 orchestrate-multimodal-auditor

Your own site · 80×15
<a href="https://agentmods.dev/skills/nitish-r-g/hackerrank-orchestrate-skills/orchestrate-multimodal-auditor"><img src="https://agentmods.dev/badge/skills/nitish-r-g/hackerrank-orchestrate-skills/orchestrate-multimodal-auditor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 55 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 701 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 $0.00055 $0.00701
Opus 5 $0.00028 $0.00351
Sonnet 5 $0.00011 $0.00140
Haiku 4.5 $0.00006 $0.00070

Measured 12d ago against content hash 3c6a3a41a16f, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

orchestrate-multimodal-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 12d 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.

skills/orchestrate-multimodal-auditor/SKILL.md · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Orchestrate: Multimodal Auditor

Evidence tier: first-hand build (August 2026). Grounded in a completed Orchestrate submission that was audited to destruction — 48 logged defects, 9 measured-and-rejected optimisations, 17 certification scripts. Every number below was measured on that system. Nothing here claims access to HackerRank's internal scoring.

The rule

Calling an OCR or ASR API is not multimodal reasoning. Media must materially change the final decision, and you prove it by turning the modality OFF.

To claim modality X drives outcome Y, show both:

  1. with X enabled, Y happens, and
  2. with X disabled, Y stops happening.

Only (1) is compatible with X being decorative.

What this caught in a real build

Two findings, opposite directions:

The pipeline was dark. Hosted vision returned 0 characters on 15/15 image rows, and that was written down as "the images carry no signal". Wrong — it was a statement about two remote services, not the corpus. A local engine extracted 10,104 characters from 19 of 20 images.

The proof it was live. Generated posters whose scam text existed only in pixels, with innocuous accompanying message text:

                  OCR ON        OCR OFF
otp harvest       mute/scam     digest/unknown
wallet + shortlink mute/scam    digest/unknown
...
                  7/7 caught    7/7 escaped

Both halves hold, so image content determines the verdict.

Do not let media reclassify intent

Measured: feeding OCR text into intent lexicons regressed two rows — an unrelated calendar screenshot and a generic brochure flipped them from urgent to event.

The scoping rule that fixed it, derived from the spec itself: message_text is empty on 8/8 voice rows (so the transcript is the message and participates fully) and non-empty on 15/15 image rows (so the poster is a supplementary artifact — it may escalate a safety signal, but must not redefine what the sender meant).

The checklist

Read the full file on GitHub · 68 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. 12d ago First seen · 68 lines · 0 tokens per session scan A 3c6a3a41a16f

Subscribe to this mod's changes

orchestrate-multimodal-auditor is a skill published in the GitHub repository NITISH-R-G/hackerrank-orchestrate-skills (3 stars, last pushed 1mo ago), licensed MIT. It adds 55 tokens to every session and 701 once invoked, about $0.0003 per session on Opus 5. 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.