Parser-only instruction found in non-visible HTML block.
Local-first scanner plus MCP guardrail
Catch hidden instructions before they reach your model.
Veridicus Scan inspects HTTPS URLs plus PDF, DOCX, and HTML files for hidden prompt-injection patterns, suspicious metadata, parser-visible drift, and risky redirect behavior. Premium also unlocks a local MCP mode for App Intents and agent workflows while the app is active.
- Inputs
- HTTPS URLs plus imported PDF, DOCX, and HTML files
- Checks
- Hidden channels, metadata anomalies, parser-visible differences, hidden Unicode, and URL redirect policy
- Outputs
- Risk score, risk band, findings, guidance, coverage notes, and PDF or JSON export
scan.session
LOCAL
Veridicus Scan
Hidden instruction inspection report
Risk score
84
Directive-style language detected in a document metadata field.
Redirect blocked before the destination fetch continued in strict URL mode.
report.export = pdf,json
source.trust = direct
What matters
Show the scan as a calm evidence system, not a noisy dashboard. Trust comes from readable proof.
Start here
Choose the product view you need.
Jump straight to the part of Veridicus Scan that matches your decision: why hidden instructions matter, what the app covers, and which workflows it supports.
Why it matters
See the gap between what a person sees, what parsers normalize, and why hidden instructions matter before content reaches AI.
Coverage
Review the exact source types, hidden channels, report outputs, and partial-coverage notes the app actually supports today.
Use cases
Go from headline messaging into real scenarios for assistant safety, document review, evidence export, and MCP automation.
OpenClaw + Veridicus Scan
Use Veridicus Scan before OpenClaw reads the internet.
If OpenClaw will fetch URLs, read files, or run tool-enabled workflows, put Veridicus Scan in front of it. Set up the agent, scan suspicious pages and documents locally, then pass only reviewed content into the workflow.
How to set up OpenClaw on macOS
Install the CLI, run onboarding, pair a channel, open the dashboard, and run the basic doctor and security checks.
Use Veridicus Scan as the OpenClaw safety layer
Screen suspicious URLs, PDFs, DOCX files, and HTML locally before OpenClaw fetches, uploads, or reasons over them.
Add a local MCP scan step
Keep the intake decision close to the agent boundary when you want repeatable local automation instead of blind ingestion.
Workflow
Scan like an evidence workflow, not a black box.
The app already follows a real sequence users can understand: choose a source, analyze normalized content and hidden channels, then review a scored report with guidance.
Choose the source
Start from an HTTPS URL or import a PDF, DOCX, or HTML file. The app keeps the source type explicit from the first screen.
Inspect visible and hidden channels
Normalize content, compare parser-visible and visible text where possible, inspect metadata and hidden-channel artifacts, and enforce the chosen URL mode.
Export a report you can act on
Every scan ends with a risk score, risk band, findings, guidance, and coverage notes, with PDF or JSON export available from the share screen.
Premium MCP mode
Use Veridicus Scan as a local MCP guardrail for agent workflows.
MCP is not an afterthought in the app. Premium unlocks a local, foreground, session-based bridge for App Intents and external local wrappers so an agent can scan text or URLs, fetch reports, redact memory, and gate risky plans without moving the workflow to a remote daemon first.
Core MCP scan path
Open Session, Scan Text, Scan URL, Get Report, Export Report, and Close Session cover the main local scan-and-report loop.
Memory and disclosure controls
Ingest Memory, Retrieve Memory, Selective Disclosure, and Evaluate Selective Disclosure support tighter agent-runtime handling.
Plan and action guardrails
Scope Tools, Guard Plan, and Gate Action let the app participate in local policy enforcement instead of only returning a scan score.
Report-first trust
Every scan should end in proof you can read quickly.
The report is not a side effect. It carries the score, band, findings, coverage state, and export controls that users actually rely on after a scan.
REPORT / FINDINGS / LOCAL EXPORT
source
https://example.com/help-center
finding
parser-only instruction found in hidden DOM
evidence
non-visible block contained directive-style control language
action
review before sending into an assistant or sharing into an AI workflow
Trust model
Local-first by default. Clear boundaries when the network matters.
Privacy and scan boundaries should be stated with precision. Users should understand what stays on-device, when a user-provided URL is fetched, and that exported evidence can be shared with snippets redacted by default.
On-device inspection
Lead with local analysis and local report generation instead of uploading content first and explaining the risk later.
Explicit scan boundaries
Use exact language around HTTPS-only URL entry, strict or lenient redirect handling, and partial-coverage notices.
Readable evidence
Make findings, guidance, and coverage notes understandable enough that users can defend the decision they make after the scan.
FAQ
Short answers, written clearly.
What does Veridicus Scan inspect?
Today it is built around HTTPS URL scans and imported PDF, DOCX, and HTML files, with analysis aimed at hidden prompt-injection patterns, metadata anomalies, and parser-visible signals users may not see directly.
How do URL scans handle redirects?
URL input is HTTPS-only. Strict mode blocks cross-URL redirects, while lenient mode follows HTTPS-safe redirects for broader compatibility.
What can I export after a scan?
The share flow supports JSON and PDF export. Evidence snippets are redacted by default unless the user explicitly chooses to include them.
What is the MCP feature?
Premium unlocks a local, foreground MCP and App Intent bridge for session-based agent workflows. It supports Scan Text, Scan URL, Get Report, Export Report, plus runtime-defense methods such as Selective Disclosure, Guard Plan, and Gate Action while the app is active.
Why would a report mention partial coverage?
If a fetched response or imported file exceeds the configured scan budget, the app marks the report as partial and includes coverage notes so the result is not overstated.
Final pass
Inspect the content before the model does.
Veridicus Scan is built for people who want local inspection, readable evidence, and clean boundaries before content enters an AI workflow.