ClaudeMark: Skill for Claude Code

.agents/skills/ai-forensics/SKILL.md

ai-forensics is a skill for Claude Code, Codex from karthikrshet/ClaudeMark. It costs 28 tokens per session (484 once invoked), scanned A, original, MIT.

A local tool for examining possible AI watermarks, hidden Unicode characters, steganography, file metadata, and content-provenance records. Provenance records describe where a file came from and how it was changed.

In plain words
What is it for?
Use it to inspect text for statistical watermark patterns, find or clean zero-width characters, rewrite text for best-effort watermark disruption, and inspect or sanitize supported document and image files.
Why use it?
It helps reveal hidden markers and remove metadata before sharing or processing files you own or are authorized to handle.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: installed under .agents/ (shared by several agents).

This is karthikrshet/ClaudeMark's own configuration. It tells Claude Code and Codex how to work on ClaudeMark 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 ClaudeMark configures →

Reuse

Borrowing it

Nothing to install: this file belongs to karthikrshet/ClaudeMark. 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/karthikrshet/ClaudeMark/main/.agents/skills/ai-forensics/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/karthikrshet/ClaudeMark

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 ai-forensics

README.md
[![agentmods](https://agentmods.dev/badge/skills/karthikrshet/claudemark/ai-forensics/github.svg)](https://agentmods.dev/skills/karthikrshet/claudemark/ai-forensics)
Your own site
<a href="https://agentmods.dev/skills/karthikrshet/claudemark/ai-forensics"><img src="https://agentmods.dev/badge/skills/karthikrshet/claudemark/ai-forensics/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 ai-forensics

Your own site · 80×15
<a href="https://agentmods.dev/skills/karthikrshet/claudemark/ai-forensics"><img src="https://agentmods.dev/badge/skills/karthikrshet/claudemark/ai-forensics.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 28 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 484 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 $0.00028 $0.00484
Opus 5 $0.00014 $0.00242
Sonnet 5 $0.00006 $0.00097
Haiku 4.5 $0.00003 $0.00048

Measured 10d ago against content hash 51ea0eba9ca3, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

ai-forensics 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 10d 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.

.agents/skills/ai-forensics/SKILL.md · 52 lines

What it actually says

AI Forensics & Watermark Analysis Skill

Use this skill to inspect, analyze, disrupt, and sanitize AI watermarks and provenance metadata from text, documents, and images on content the user owns or is authorized to process.

Available Actions

1. Statistical Watermark Analysis

Analyze text for subtle generative sampling biases and statistical regularities:

python claudemark.py analyze "<text>" --algorithm claude

2. Unicode & Steganography Inspection

Inspect and visualize hidden zero-width characters (<ZWSP>, <BOM>, <RLO>):

python claudemark.py unicode inspect file.txt
python claudemark.py unicode visualize file.txt
python claudemark.py unicode clean file.txt -o clean.txt

3. Best-Effort Statistical Watermark Disruption

Restructure cadence and rotate high-frequency transition tokens:

python claudemark.py rewrite file.txt -o rewritten.txt --strategy synonym_cadence
python claudemark.py evaluate file.txt rewritten.txt

4. File Provenance & Document Cleaning

Scrub metadata from 10 document and image formats (PDF, DOCX, ODT, HTML, MD, TXT, PNG, JPEG, WebP, SVG):

python claudemark.py inspect document.pdf
python claudemark.py clean document.pdf -o clean.pdf

5. Defensive Security Scanner

Scan files for zip bombs, malicious PDF actions, macros, and path traversal:

python claudemark.py security scan upload.pdf

REST API Integration

If the server is running on http://127.0.0.1:8765:

  • Analyze Text: POST /api/analyze {"text": "..."}
  • Inspect Unicode: POST /api/unicode/analyze {"text": "..."}
  • Clean File: POST /clean {"file": "<base64>", "name": "doc.pdf"}
  • Agent Tool Dispatch: POST /api/agent/exec {"tool_name": "...", "arguments": {...}}
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. 10d ago First seen · 52 lines · 28 tokens per session scan A 51ea0eba9ca3

Subscribe to this mod's changes

ai-forensics is a skill published in the GitHub repository karthikrshet/ClaudeMark (9 stars, last pushed 16d ago), licensed MIT. It adds 28 tokens to every session and 484 once invoked, about $0.0001 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.

Related

Other skills, from other repositories

file-to-markdown

Turn almost any file into Markdown plus metadata — PDF, Office, HTML, email, archives, images, audio/video, 1000+ formats — powered by Apache Tika, either via the tika-app CLI (zero setup, one file) or a running tika-server (curl, warm process, many calls). Leads with rmeta (structured, embedded-item-aware output) as…

apache/tika · 147 tokens

file-forensics

Examine what a file claims about itself and what it actually contains — powered by Apache Tika. True content-based type detection (extensions lie), provenance claims (authors, dates, creating application), revision and tamper signals (PDF incremental updates, tracked changes, hidden slides, zip integrity), hidden and…

apache/tika · 142 tokens

gdocs-audit-report

Write and format security-audit reports in Google Docs via the Docs API - findings, tables, and severity styling. Use to build or fix a report.

forefy/.context · 37 tokens

employment-contract-templates

Create employment contracts, offer letters, and HR policy documents following legal best practices. Use when drafting employment agreements, creating HR policies, or standardizing employment documentation.

aisa-group/skill-inject · 37 tokens

literature

Load when extracting GEO accessions, dataset metadata, and downloadable references from a scientific paper (PDF / URL / DOI / PubMed ID / raw text) for downstream omics analysis. Skip when the dataset is already in hand; only routing a query (use orchestrator).

TianGzlab/OmicsClaw · 58 tokens

cms-metadata-tagger

Takes a finished piece of content and produces a complete, archive-ready metadata record — including descriptive tags, subject classifications, rights notes, and an archival summary — for entry into a content management or digital archive system.

ur-grue/autopunk-media-skills · 49 tokens