Bound approvals
One-time decisions are consumed before dispatch; expired, duplicate or mismatched completion receipts are rejected. Current policy and page identity are rechecked.
Agent Guard 1.4.0 · Published
AI Agent Firewall
Review stronger approval checks, inspect local diagnostics and end tasks explicitly.
Chrome Web Store · 1.4.0
Separate codebase Policy-first Auditable
Prompt & context
Agent Guard policy gateway Allow Deny / redact
AI agent
Tools / MCP
Outbound data
Human approval
What’s new · Published
Agent Guard 1.4.0
Review stronger approval checks, inspect local diagnostics and end tasks explicitly.
Diagnostic Center
Check collector reachability and local task/bridge state. Export excludes hosts, URLs, identifiers, page text and audit content; reachability is not proof of full protection.
Task lifecycle
View expiry and confirm End task & lock. In-flight actions and page changes prevent unsafe task edits. Already-sent actions cannot be recalled.
Reliable local context
Fix bridge summary handling, reject stale responses and preserve bounded audit snapshots. Existing permissions and local-only operation remain unchanged.
These features operate on supported browser-visible evidence. No additional extension permissions or extension analytics were added.
Chrome Web Store · 1.4.0
Separate product
View the SiteEgress product
Agent Guard is not the SiteEgress Chrome extension.
SiteEgress inspects browser-visible website privacy behavior. Agent Guard supervises supported browser-visible agent actions on sites the user authorizes. The products have separate codebases, permission models, interfaces, privacy policies, and release paths.
SiteEgress Security Suite
Two extensions. One optional local security context.
SiteEgress and Agent Guard remain independent Chrome extensions. When both are installed and both local bridge toggles are enabled, they can exchange a minimized, origin-scoped security summary on the device.
SiteEgress 1.4.0PublishedBrowser privacy & data-flow inspector
↔Local bridgeDual opt-in
Agent Guard 1.4.0PublishedBrowser-agent action security
Install independently
Use either extension alone, or install both from their separate Chrome Web Store listings.
Enable both sides
The bridge is off by default. Each extension must opt in before product summaries are exchanged.
Review shared context
Agent Guard can see SiteEgress destination risk context; SiteEgress can show minimized Agent Guard action and policy context.
Minimized fields only
- HTTP(S) origin or destination hostname
- Known service or action category
- Aggregate scores, counts, policy outcomes, and confidence
- Versioned evidence metadata with bounded collections
Never exchanged
- Page text or prompts
- Typed form values, passwords, secrets, Cookie or Storage values
- Request or response bodies and URL parameter values
- Raw destination IP addresses
Bridge context is evidence, not authority. SiteEgress cannot approve an Agent Guard action, and Agent Guard cannot rewrite a SiteEgress audit.
SiteEgress · Agent Guard
See the change. Understand the next step.
Explore service context, a baseline comparison and a reviewable result. No installation or personal browsing data required.
Try a sample report
Intended capabilities
Runtime controls designed for agentic systems
These capabilities describe the published release. Coverage remains limited to supported browser-visible channels and is not complete protection against every Agent or prompt-injection risk.
Prompt and content risk screening
Identify prompt injection patterns, untrusted instructions, hidden directives, and risky content before an agent acts.
Tool and MCP policy
Define which tools, MCP servers, methods, and resource scopes an agent may use for a given task or identity.
Data egress controls
Inspect and constrain what information may leave trusted boundaries through tools, APIs, messages, or generated files.
Human approval gates
Require explicit review for high-impact actions such as payments, destructive changes, external communications, or secret access.
Secrets and PII safeguards
Detect sensitive values and apply masking, denial, substitution, or approval policies before tool execution.
Auditable decision traces
Record policy inputs, decisions, approvals, tool calls, and outcomes in a reviewable security trail.
Product architecture
A policy gateway between intent and action
Agent Guard is designed as a control point around an agent—not as another general-purpose chatbot. Policies should be testable, decisions explainable, and failures visible.
Testable policies
Explainable decisions
Visible failures
Prompt & context
Agent Guard policy gateway Allow Deny / redact
AI agent
Tools / MCP
Outbound data
Human approval
Designed for
Who Agent Guard is for
Agent platform teams
Add a consistent policy layer across internal agents, tools, and MCP integrations.
Security and governance
Review high-risk actions, data flows, exceptions, and policy coverage.
SaaS builders
Offer customer-facing agents with clear scopes, approvals, and auditability.
Automation teams
Protect workflows that touch email, finance, support systems, code, or production infrastructure.
Available on the Chrome Web Store
Add Agent Guard to Chrome
Install Agent Guard
Install Agent Guard from its official listing. Publication does not imply complete channel coverage, automatic compliance, or a guarantee against every Agent threat.
Add practical guardrails to supported AI Agent actions
Both extensions are available on the Chrome Web Store. Install them independently from their official listings. SiteEgress 1.4.0 · Agent Guard 1.4.0