Repo Risk Score
A clear executive number for how exposed your codebase is to AI-agent mistakes.
Feature library
Every CapitalGuard feature is built around AI-agent repository exposure.
The product combines scanner checks, firewall-style guardrails, delivery workflow, reporting, verification, and partner-ready assets into one commercial license.
Feature stack
Scanner to proof workflow
DetectFind dangerous files, prompt-injection paths, unsafe workflow automation, exposed secret context, and agent tool bridges.
PreventGenerate policy starters, protected paths, redaction rules, command limits, and human-review checkpoints.
DeliverMove the buyer through license activation, authorized scope intake, report handoff, support recovery, and proof verification.
9
Exposure modules
3
License packages
10
Pro scan credits
1 year
Updates included
Core features
The feature set is focused, not bloated.
CapitalGuard sells what a buyer actually needs to start: risk detection, a prevention packet, and a clean handoff from payment to authorized scanning.
Dangerous File Detection
Flags files agents should not read, edit, summarize, or send to third-party tools.
Secret Exposure Scan
Finds risky secret patterns and always redacts sensitive values in the UI.
Unsafe Permission Review
Maps what AI tools can inspect, run, upload, edit, and connect to.
Prompt-Injection Report
Shows hidden instruction paths inside code, docs, issues, prompts, and logs.
Policy Generator
Generates firewall-style Codex, Claude, Cursor, and Copilot guardrails your team can install.
Monitoring Dashboard
Track risk changes across repos, seats, tools, permissions, and policies.
Executive PDF Report
Board-ready summaries with business impact, confidence, and next actions.
Trust Badge
Paid teams can publish a restrained, verifiable CapitalGuard scanned badge.
Before and after
Turn vague AI-agent concern into a documented operating model.
The scanner does not claim to fix everything. It gives the buyer a clear sequence for detecting, preventing, and communicating repository exposure.
Before
AI tools can inspect broad repository context by default.
After CapitalGuard
Sensitive paths, dangerous files, prompts, workflows, and secret patterns are mapped before wider agent use.
Before
Security checks are scattered across tools, docs, and manual reviews.
After CapitalGuard
CapitalGuard packages scan intake, report delivery, policy files, and buyer handoff in one license path.
Before
Teams talk about AI risk without a concrete artifact.
After CapitalGuard
Buyers get a redacted report, local guardrails, certificate path, and executive-ready evidence.
Product ecosystem
Features connect to delivery, not just a marketing checklist.
Each module has a place in the sales path: buyer confidence before checkout, useful artifacts after payment, and a controlled claim boundary for partners.
Scanner
Detect dangerous files, secret exposure patterns, unsafe permissions, and prompt-injection paths.
Firewall Installer
Give the buyer local guardrail files they can install into the repository workflow after the report.
Delivery Vault
Keep license activation, buyer kit, scan intake, certificates, and access recovery in one place.
Partner Engine
Approved creators can promote the product with tracked links, payout rules, and proof assets.
Package mapping
Pick the license by scan capacity and evidence depth.
Fastest start
Starter License
€49
Find the first AI-agent exposure risks in one small repository.
AI-agent risk score
Dangerous files summary
Exposed-secret risk summary
PDF report
Best founder value
Growth License
€149
Turn AI-agent uncertainty into a founder-ready security packet.
Everything in Starter
Prompt-injection exposure report
AI coding tool safe-use checklist
Policy file generator
Most complete
Pro License
€349
Protect the main product repo before AI agents get more authority.
GitHub integration path
Full dashboard
Prevention checklist
Downloadable executive report
Operational surfaces
The same features are reflected in the buyer dashboard.
These public dashboard metrics are demo values that show the product shape without publishing private customer data.
AI Agent Risk Score
78/100
High exposure
Dangerous files
7
Require policy blocks
Exposed secret risks
4
Redacted in report
Unsafe permissions
5
Tool access review
Prompt-injection paths
3
Docs and issues
Policy file
1
Ready to deploy
Ready path