Free tools Windows power users keep installed
One-click scans. No signup required.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Revoke an AI agent as an identity incident, not merely a stopped process. Pause the runtime, preserve logs, block the agent’s workload identity at its issuer, revoke every delegated OAuth/OIDC grant and tool connection, rotate persistent secrets, and investigate use before rebuilding. Deleting one API key can leave short-lived tokens issued from the same service identity working.
The objective is a verified cutoff with evidence intact, followed by a least-privilege rebuild that uses a unique workload identity and short-lived, audience-restricted credentials.
What to do first
Use this order when an agent may be compromised. The sequence separates stopping new activity from invalidating credentials that could be used by another process.
- Declare and scope the incident. Record the agent runtime or orchestration instance, owner, workload or service account, API keys, OAuth and OIDC grants, tool or MCP connections, network routes, repositories, data stores, CI/CD jobs, and other agents that share any of those identities.
- Preserve evidence. Export authentication, service-account, key-use, tool-call, network-egress, repository, and data-access logs. Keep timestamps in a common time zone, record the export method, and retain a controlled copy. Google Cloud guidance and the Australian Cyber and Infrastructure Security Centre both emphasize reviewing affected resources and unfamiliar runtimes, sessions, service accounts, identities, and logs before restoration.
- Stop new actions. Pause the workflow, disable its scheduler or webhook, quarantine the worker, and block egress to sensitive tools and data services. If possible, keep a forensic copy of the runtime rather than destroying it.
- Block the identity at the issuer. Disable or delete the service account, or apply an IAM deny policy that blocks the principal while the investigation continues. This is the step that can stop active tokens issued for that identity; deleting only a key does not reliably do so.
- Remove delegated access. Revoke OAuth grants, SSO authorizations, personal access tokens, deploy keys, SSH keys, and tool-server credentials at every connected provider. Perform the action separately in each service; revoking a grant in one provider does not remove grants elsewhere.
- Rotate persistent secrets. Create replacement keys only after the compromised path is blocked. Install each replacement in its legitimate consumer, test it, then disable and delete the old secret. Remove roles that allow an unauthorized principal to mint or impersonate tokens, including the Google Service Account Token Creator role where it is present.
- Investigate before rebuilding. Determine whether the agent read, changed, or exfiltrated data; called tools; modified repositories or deployment definitions; or created additional credentials.
- Restore with reduced authority. Give the replacement agent a distinct workload identity, narrowly scoped and short-lived credentials, audience restrictions, and monitoring. Keep the old identity blocked until the investigation and recovery review are complete.
Why stopping the model is not enough
An agent usually acts through several layers: a model runtime, an orchestration worker, a service identity, API keys, delegated OAuth grants, tool-server credentials, CI/CD secrets, and permissions in downstream systems. Stopping the model process prevents that process from making new calls, but it does not invalidate a token copied from memory, a key stored in a build system, or a grant held by a separate integration.
#1 Best Overall
- POWERFUL SECURITY KEY: The YubiKey 5C NFC is the most versatile physical passkey, protecting your digital life from phishing attacks. It ensures only you can access your accounts
- WORKS WITH 1000+ ACCOUNTS: Compatible with popular accounts like Google, Microsoft, and Apple. A single YubiKey 5C NFC secures 100+ of your favorite accounts, including email, password managers, and more
- FAST & CONVENIENT LOGIN: Plug in your YubiKey 5C NFC via USB and tap it, or tap it against your phone (NFC), to authenticate. No batteries, no internet connection, and no extra fees required
- MOST SECURE PASSKEY: Supports FIDO2/WebAuthn, FIDO U2F, Yubico OTP, OATH-TOTP/HOTP, Smart card (PIV), and OpenPGP. That means it’s versatile, working almost anywhere you need it
- PRIMARY & SPARE KEYS: Just like having a spare house key, we recommend buying two YubiKeys - one for daily use and one as a spare. That way you’ll never get locked out of your accounts
Think in terms of principals and authenticators. The principal is the service account, workload identity, or other account recognized by an issuer. Authenticators include static keys, bearer tokens, refresh tokens, certificates, SSH keys, and delegated grants. A complete response blocks the principal when necessary and invalidates each authenticator at the system that issued it.
Map every identity before changing anything
Create a one-page identity map. It gives responders a bounded target and prevents an apparently successful key deletion from leaving a second route open.
| Item to record | Questions to answer | Why it matters |
|---|---|---|
| Agent and runtime | Which model workflow, worker, container, function, or orchestration session was running? Which scheduler, webhook, or queue can start it again? | A disabled process can be relaunched by an unattended trigger. |
| Workload identity | Which service account, role, managed identity, SPIFFE identity, or other principal does it use? | Issuer-side disablement or denial is the fastest broad cutoff. |
| Persistent credentials | Which API keys, access keys, certificates, SSH keys, deploy keys, and secrets exist, where are they stored, and who consumes them? | Every copy must be replaced or invalidated. |
| Short-lived credentials | Which access, identity, or refresh tokens may already have been issued? What are their issuer, audience, expiry, and scopes? | Deleting a source key may not invalidate these tokens. |
| Delegated access | Which OAuth/OIDC grants, SSO sessions, personal tokens, and consented applications are connected? | Delegation is held at the provider and must be revoked there. |
| Tools and data | Which MCP or tool servers, repositories, databases, object stores, queues, and SaaS systems can the agent reach? | Separate tool credentials and data permissions can outlive the agent. |
| Automation paths | Which CI/CD variables, deployment roles, webhooks, scheduled jobs, and token-minting or impersonation roles can recreate access? | An attacker may regain access without the original key. |
How to revoke each credential class
Service accounts and workload identities
Use the cloud or identity provider that represents the agent’s principal. Disable the service account, delete it when policy permits, or attach an explicit deny policy that blocks its permissions while evidence is collected. Removing role bindings alone may not stop every already-issued token if another binding or trust path remains.
Check for impersonation and token-minting privileges. In Google Cloud, the Service Account Token Creator role is specifically relevant: remove it from any unauthorized principal, service account, build job, or group. Also inspect federation and trust policies that allow an external workload to assume the identity.
Static API keys and access keys
For each key, record its identifier, owner, creation date, last-use time, scopes, and consuming applications. Disable or revoke the old key, create a replacement with only the required permissions, deploy it to the legitimate consumer, test the consumer, and then delete the old key if retention is not required for the investigation.
Do not create replacements before the compromised route is blocked; a new key can be exposed through the same runtime, secret store, or deployment job. If a key was embedded in source code or an image, assume every copy is compromised and remove it from history and artifacts according to the provider’s incident procedure.
OAuth and OIDC grants
Open the authorization or connected-app controls at each provider and revoke the grant associated with the agent. Revoke refresh tokens and active sessions where the provider offers separate controls. Narrowing a scope after the fact does not reliably undo a token that has already been issued.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchRank #2
- POWERFUL SECURITY KEY: The Security Key C NFC is the essential physical passkey for protecting your digital life from phishing attacks. It ensures only you can access your accounts.
- WORKS WITH 1000+ ACCOUNTS: Compatible with Google, Microsoft, and Apple. A single Security Key C NFC secures 100 of your favorite accounts, including email, password managers, and more.
- FAST & CONVENIENT LOGIN: Plug in your Security Key C NFC via USB-C and tap it, or tap it against your phone (NFC) to authenticate. No batteries, no internet connection, and no extra fees required.
- TRUSTED PASSKEY TECHNOLOGY: Uses the latest passkey standards (FIDO2/WebAuthn & FIDO U2F) but does not support One-Time Passwords. For complex needs, check out the YubiKey 5 Series.
- BUILT TO LAST: Made from tough, waterproof, and crush-resistant materials. Manufactured in Sweden and programmed in the USA with the highest security standards.
Repeat the review for every account used by the workflow. An agent may have one grant for a repository, another for a ticketing system, and a third for a data platform. A single sign-out action in the orchestration console does not remove those provider-side authorizations.
Tool and MCP connections
Disable the agent’s connection at the tool server and rotate the tool credential. Remove the agent’s identity from tool allow-lists, revoke server-issued sessions, and inspect tool audit logs for calls made before and after the suspected compromise. Block network egress to the tool endpoint until a replacement connection has been tested.
Separate read, write, administrative, and credential-minting permissions. An agent that only summarizes documents should not retain a tool permission that can alter production systems or create new users.
CI/CD, repositories, and deployment credentials
Inspect build variables, deployment environments, repository secrets, deploy keys, SSH keys, and automation tokens. Revoke or rotate each item at its issuer, then review workflow definitions and recent commits for unauthorized changes. A compromised agent may persist by adding a new secret, webhook, scheduled job, or deployment step even after its original key is gone.
Recommended Free Tools
Use the repository provider’s bulk and targeted revocation features where available, but verify the result with an inventory of remaining keys and tokens. The control should be auditable: record who revoked each credential, when, and which replacement was installed.
Human and administrator credentials
Protect the operator account used for containment and recovery separately from the agent. Reset or revoke it if the agent could access it, remove unnecessary administrator sessions, and require phishing-resistant authentication. A FIDO2 security key is an appropriate hardware authenticator for the administrator; it protects that human account but does not revoke agent tokens by itself.
What “revoked” actually means
| Action | What it stops | What may remain | Use when |
|---|---|---|---|
| Stop or quarantine the runtime | New calls from that process or worker | Copied keys, tokens, grants, and other runtimes remain usable | You need an immediate pause while preserving evidence. |
| Delete a persistent key | Future authentication with that specific key | Short-lived tokens already issued from the same service identity may remain valid | You are retiring one key after the identity path is contained. |
| Disable, deny, or delete the service identity | The principal’s access, including the control used to stop active tokens in the Google Cloud guidance | Independent identities and provider-side grants still need separate action | Immediate broad cutoff is required. |
| Revoke an OAuth or SSO grant | Delegated access through that provider | Grants at other providers and unrelated credentials remain | The agent accessed a connected SaaS or identity provider. |
| Invalidate an authenticator | The specific authenticator’s ability to authenticate | Other authenticators for the same account may remain | Any authenticator is suspected to be compromised. |
NIST SP 800-63B states: “CSPs SHALL promptly invalidate authenticators when … the authenticator is compromised.” Apply that rule to keys, tokens, certificates, and other authenticators, while remembering that each issuer controls its own invalidation mechanism.
Rank #3
- POWERFUL SECURITY KEY: The YubiKey 5 NFC is the most versatile physical passkey, protecting your digital life from phishing attacks. It ensures only you can access your accounts
- WORKS WITH 1000+ ACCOUNTS: Compatible with popular accounts like Google, Microsoft, and Apple. A single YubiKey 5 NFC secures 100+ of your favorite accounts, including email, password managers, and more
- FAST & CONVENIENT LOGIN: Plug in your YubiKey 5 NFC via USB and tap it, or tap it against your phone (NFC), to authenticate. No batteries, no internet connection, and no extra fees required
- MOST SECURE PASSKEY: Supports FIDO2/WebAuthn, FIDO U2F, Yubico OTP, OATH-TOTP/HOTP, Smart card (PIV), and OpenPGP. That means it’s versatile, working almost anywhere you need it
- PRIMARY & SPARE KEYS: Just like having a spare house key, we recommend buying two YubiKeys - one for daily use and one as a spare. That way you’ll never get locked out of your accounts
Verify that the cutoff worked
Do not treat a successful console message as proof of containment. Verification should test both the intended path and likely alternate paths.
- Attempt authentication with the revoked key in a controlled test and confirm rejection.
- Check that the disabled or denied service identity cannot obtain a new token or call a protected resource.
- Search logs for the identity, key identifiers, token subject, client ID, source IPs, user agents, and tool names after the revocation timestamp.
- Look for token use from another runtime, build worker, region, account, or network path.
- Confirm that schedulers, webhooks, queues, and deployment jobs cannot restart the old agent.
- Review token-minting, impersonation, federation, and secret-manager audit events for attempts to recreate access.
- Check downstream systems for newly created users, keys, grants, webhooks, repository changes, data exports, and altered retention or logging settings.
If any revoked credential still succeeds, treat the incident as active. Recheck whether the test reached the correct issuer, whether a different credential was used, whether a cached session or refresh token exists, and whether the request was authorized through another principal.
Investigate before rebuilding
Establish the timeline
Build a timeline from the first suspicious model output, tool call, login, key use, configuration change, or unusual egress event through containment. Normalize timestamps and preserve original log files. Record both successful and denied requests; denied requests can reveal attempted persistence.
Determine what the agent could do
For each permission, distinguish reading, writing, deleting, invoking tools, changing configuration, and minting or impersonating credentials. Then compare those capabilities with observed calls. A token that was never used to write production data has a different recovery decision from one that changed deployment definitions.
Assess data exposure
Inspect repository access, database queries, object-store reads, downloads, exports, and network destinations. Identify the data classes involved and retain hashes or manifests of relevant artifacts where your evidence policy permits. Do not rebuild over the original environment until required evidence is captured.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Record the decision
Document the suspected entry point, identities revoked, systems checked, evidence retained, confirmed actions, unknowns, and the person approving restoration. This record supports notification decisions and prevents a later responder from re-enabling an unreviewed credential.
Rebuild the agent with safer credentials
Restore only after the old identity and its delegated access are controlled. The replacement should be materially different, not a copy of the compromised configuration.
Rank #4
- POWERFUL SECURITY KEY: The Security Key NFC is the essential physical passkey for protecting your digital life from phishing attacks. It ensures only you can access your accounts.
- WORKS WITH 1000+ ACCOUNTS: Compatible with Google, Microsoft, and Apple. A single Security Key NFC secures 100 of your favorite accounts, including email, password managers, and more.
- FAST & CONVENIENT LOGIN: Plug in your Security Key NFC via USB-A and tap it, or tap it against your phone (NFC) to authenticate. No batteries, no internet connection, and no extra fees required.
- TRUSTED PASSKEY TECHNOLOGY: Uses the latest passkey standards (FIDO2/WebAuthn & FIDO U2F) but does not support One-Time Passwords. For complex needs, check out the YubiKey 5 Series.
- BUILT TO LAST: Made from tough, waterproof, and crush-resistant materials. Manufactured in Sweden and programmed in the USA with the highest security standards.
- Create a unique workload identity. Do not reuse a human account or share one broad service identity across agents.
- Assign least privilege. Separate permissions for reading data, writing data, invoking tools, deploying code, and minting credentials. Grant only the operations required for the specific workflow.
- Prefer short-lived credentials. NIST describes long-lived API keys and bearer tokens as broad bearer credentials. Use dynamically issued credentials with narrow scopes, short expiry, and an audience limited to the intended service.
- Use standard, verifiable mechanisms. NIST SP 800-228 recommends cryptographic verification, regular credential cycling, signing-key rotation, and standard protocols such as OAuth2/OIDC, SPIFFE SVIDs, and verifiable JWTs.
- Constrain the network. Allow egress only to named tool and data endpoints. Deny direct access to administrative interfaces and metadata services unless the workflow demonstrably requires it.
- Validate every token. Check issuer, signature, audience, subject, expiry, not-before time, and required scopes. Do not accept a bearer token merely because it is structurally valid.
- Deploy in stages. Start with read-only or synthetic data, observe calls, then enable the minimum write operations. Keep a fast rollback path that does not depend on the agent’s own credentials.
Build a tested kill switch
A kill switch should be an operator-controlled action that blocks the workload identity and tool egress without waiting for the model or its orchestration code. Test it regularly in a non-production environment and measure the time from activation to denied calls.
- Identity control: disable or deny the agent’s workload identity.
- Network control: block outbound connections to tool and data endpoints.
- Trigger control: disable schedules, webhooks, queues, and deployment jobs.
- Secret control: revoke the agent’s persistent keys and refresh tokens.
- Evidence control: preserve logs and retain a read-only forensic copy.
- Approval control: require a separate, phishing-resistant administrator account to restore access.
Keep these controls independent. If the agent can alter the same policy or secret store used for its kill switch, the switch is not a reliable containment mechanism.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsCommon mistakes and their fixes
Deleting one key and declaring victory
Failure: An already-issued short-lived token, refresh token, OAuth grant, or second key continues to work.
Fix: Disable, deny, or delete the underlying service identity when an immediate cutoff is required, then revoke delegated access and rotate every persistent secret in the identity map.
Rebuilding before collecting logs
Failure: New deployments overwrite timestamps, volatile files, or evidence of persistence.
Fix: Export and protect logs, capture the runtime state where feasible, and write the incident timeline before replacement.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Revoking only in the orchestration console
Failure: Provider-side grants and tool credentials remain active.
Best Value
- FIDO2 & Passkey Ready: Business-ready and FIDO2 L1 certified. This key is supported by major management suites and is ideal for both individual and enterprise deployment. Works seamlessly with Gmail, Facebook, GitHub, Dropbox, Coinbase, and more.
- Universal Connectivity (USB-A ): Features a built-in USB-A connector—simply unfold the key and plug it into your compatible PC or laptop for seamless authentication on the go.
- Dedicated Manager App: Use the Thetis Manager App for the initial hardware PIN setup. Setting the PIN on the device first ensures a smooth registration process. Once the PIN is configured, you can begin registering the key across your favorite FIDO2-compatible online services.
- Ultra-Durable & Portable: Featuring a rotating metal cover, this key is water, crush, and tamper-resistant. It fits easily on a keychain and requires no batteries or network connectivity.
- Check FIDO2 compatibility before purchase - Known limitations: ID Austria is not supported (requires FIDO2 Level 2). Windows Hello login only works with Windows Enterprise editions that support Entra ID, and NFC is NOT supported.
Fix: Revoke at each issuer and connected service, including repositories, CI/CD, SaaS applications, and tool servers.
Rotating into the same exposed path
Failure: A new secret is immediately disclosed through the compromised worker, image, build variable, or repository.
Fix: Block the old path first, replace the runtime or trust relationship, and deploy the new secret only to the verified consumer.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Restoring the old permissions
Failure: The rebuilt agent retains broad write, administrative, or token-minting authority.
Fix: Create a distinct identity with narrow scopes, audience restrictions, short expiry, and separate read, write, tool, and credential-minting roles.
Control choices by response objective
| Control | Cutoff speed | Token coverage | Typical blast radius | Auditability |
|---|---|---|---|---|
| Pause runtime or worker | Very fast for that process | None outside the process | Single runtime | Usually clear in orchestration logs |
| Disable or deny workload identity | Fast at the issuer | Broad coverage for that principal, including the control needed for active-token cutoff | All workloads using the identity | Strong issuer audit trail |
| Delete or disable a persistent key | Fast for that key | That key only; not necessarily issued tokens | Consumers of the key | Key-management event is recorded |
| Revoke OAuth/OIDC grant | Depends on provider | Delegated access at one provider | One user, app, or grant | Provider authorization log |
| Block tool egress | Fast at the network boundary | Calls to the blocked endpoint | All workloads sharing the route | Firewall or gateway logs |
Use more than one control when the agent has multiple identities or when the time of compromise is uncertain. The narrowest control limits collateral damage; the issuer-side identity block provides the broadest immediate cutoff for a shared principal.
Quick Recap
Operational checklist
- Agent runtime, owner, scheduler, and network path identified.
- Service account or workload identity identified and blocked at its issuer.
- API keys, certificates, SSH keys, deploy keys, refresh tokens, and personal tokens inventoried.
- OAuth/OIDC grants and SSO sessions revoked at every connected provider.
- Tool and MCP credentials disabled and rotated.
- CI/CD variables, repository secrets, webhooks, and token-minting roles reviewed.
- Authentication, tool, network, repository, and data-access logs preserved.
- Post-revocation token use and credential-recreation attempts monitored.
- Data access and possible exfiltration assessed before restoration.
- Replacement agent deployed with a unique identity, least privilege, short-lived credentials, audience restrictions, and tested kill switch.
- Administrator account protected with phishing-resistant authentication, such as a FIDO2 security key.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
