# Web3Sec > Independent security services, practical operational guides, and source-linked incident analysis led by Raiders. This supplementary index links to canonical public pages and faithful text exports. Incident/advisory dates describe events or disclosures; publication/update dates describe these articles. The library is curated, not a live alert feed. Guides and diagrams propose controls rather than proving they exist in a reader's system. ## Start here - [Homepage](https://www.web3sec.news/) - [Services](https://www.web3sec.news/services) - [Scope an engagement](https://www.web3sec.news/engagements) - [Fictional finding specimen](https://www.web3sec.news/sample-report): Illustrative teaching example, not client work or a live test. - [About Raiders](https://www.web3sec.news/about) - [Approach](https://www.web3sec.news/approach) - [Research methodology](https://www.web3sec.news/methodology) - [Editorial policy](https://www.web3sec.news/editorial-policy) - [Contact](https://www.web3sec.news/contact) - [Privacy](https://www.web3sec.news/privacy) - [Working library](https://www.web3sec.news/blog) - [Field guides](https://www.web3sec.news/guides) - [Agent field manual](https://www.web3sec.news/guides/agents) - [Incident studies](https://www.web3sec.news/latest) - [Atom feed](https://www.web3sec.news/feed.xml) - [Sitemap](https://www.web3sec.news/sitemap.xml) ## Planning tools - [Tools index](https://www.web3sec.news/tools) - [Agent permissions planner](https://www.web3sec.news/tools/agent-permissions): Client-only worksheet for permissions, approvals, revocation and evidence; no account connection or scan. - [Research lab preflight](https://www.web3sec.news/tools/research-preflight): Client-only worksheet for source, authorization, lab boundaries, egress and disposal; no code execution. ## Services - [Smart contract security](https://www.web3sec.news/services/smart-contract-security): Challenge the assumptions, permissions, and economic edge cases in the code that moves value. - [Web & API pentesting](https://www.web3sec.news/services/web-api-pentesting): Follow the attack paths between identities, application workflows, APIs, and the systems they can reach. - [Incident readiness & response](https://www.web3sec.news/services/incident-response): Prepare the decisions, access, and evidence your team will need when normal assumptions stop holding. - [AI & supply-chain security](https://www.web3sec.news/services/ai-supply-chain-security): Examine what your agents can act on, what your build trusts, and how untrusted input can reach a privileged action. - [Technical writing & research](https://www.web3sec.news/services/technical-writing): Turn complex engineering and security work into material that readers can understand, inspect, and use. - [Security strategy & technical GTM](https://www.web3sec.news/services/security-advisory): Connect security priorities, product decisions, and a technical story that stands up to scrutiny. - [Mobile app security testing](https://www.web3sec.news/services/mobile-app-security): Follow the boundaries between the mobile app, the device, external entry points and the APIs that authorize real actions. ## Agent security guides - [AI agent and MCP security: constrain the tool, not just the prompt](https://www.web3sec.news/blog/ai-agent-mcp-security-runbook): A runbook for agent integrations, local servers, authorization, data access and realistic prompt-injection tests. [guide; analysis published 2026-11-08] - [Connect an AI assistant to email and calendars without delegating your entire identity](https://www.web3sec.news/blog/personal-ai-assistant-email-calendar-security): Decide what the assistant can read, draft, send and delete, then rehearse a malicious-message and disconnected-account scenario. [guide; analysis published 2026-11-08] - [Financial AI agents: separate research, transaction preparation and execution authority](https://www.web3sec.news/blog/financial-ai-agent-permission-boundaries): Design payment and treasury workflows around verified destinations, transaction-bound approvals, limits and an independent stop control. [guide; analysis published 2026-11-08] - [AI-assisted bug bounty work: keep scope, execution and evidence under researcher control](https://www.web3sec.news/blog/ai-assisted-bug-bounty-workflow): Use agents for understanding and reproducible reporting while preserving program authorization, data boundaries and meaningful human validation. [guide; analysis published 2026-11-08] - [Before installing an MCP server: review the package, process and connected account separately](https://www.web3sec.news/blog/mcp-server-installation-review): A preflight and change-review procedure for local and remote MCP integrations, from installation hooks to token revocation. [guide; analysis published 2026-11-08] - [Agent credentials: issue the smallest useful secret and prove you can revoke it](https://www.web3sec.news/blog/agent-secrets-scope-and-recovery): An operating procedure for task-scoped identities, secret delivery, output review and recovery after unintended access or upload. [guide; analysis published 2026-11-08] - [Browser agents and logged-in sessions: protect the account behind the tab](https://www.web3sec.news/blog/browser-agent-session-security): Review the authority inherited from cookies, browser profiles, connected sites and downloads before an agent uses an authenticated browser. [guide; analysis published 2026-11-08] - [Agent memory and RAG: prevent retrieved text from becoming persistent policy](https://www.web3sec.news/blog/agent-memory-retrieval-poisoning): Track source provenance, enforce current retrieval permissions and test whether poisoned content survives into later tasks. [guide; analysis published 2026-11-08] - [Review agent skills and plugins as instructions, dependencies and delegated authority](https://www.web3sec.news/blog/agent-skills-plugin-review): Inspect the whole instruction bundle, referenced resources and effective permissions before enabling a workflow extension or accepting its update. [guide; analysis published 2026-11-08] - [Unattended agents: bound schedules, retries, cumulative usage and stale approvals](https://www.web3sec.news/blog/unattended-agent-schedules-and-budgets): Make repeated agent work stop predictably when permissions, timing, costs or external side effects no longer match the approved task. [guide; analysis published 2026-11-08] - [Agents, localhost and internal networks: enforce what the executor can actually reach](https://www.web3sec.news/blog/agent-localhost-network-boundaries): Protect local bridges, development services, cloud metadata and internal APIs with explicit network and application boundaries. [guide; analysis published 2026-11-08] ## Other field guides - [Identity security: phishing-resistant login, usable recovery and verified offboarding](https://www.web3sec.news/blog/identity-mfa-recovery-runbook): An operating procedure for the accounts that control email, cloud infrastructure, source code and company recovery. [guide; analysis published 2026-11-08] - [Email security: protect the recovery inbox and control who can send as your domain](https://www.web3sec.news/blog/email-account-security-runbook): Separate mailbox compromise, domain spoofing and support impersonation into controls with observable tests. [guide; analysis published 2026-11-08] - [Domain and DNS security: preserve control of the address your users trust](https://www.web3sec.news/blog/domain-dns-security-runbook): A runbook for registrar ownership, scoped DNS changes, renewal resilience and safe nameserver recovery. [guide; analysis published 2026-11-08] - [Developer device security: isolate experiments from production authority](https://www.web3sec.news/blog/developer-device-security-runbook): A practical baseline for laptops that handle source code, external repositories, credentials and sensitive client material. [guide; analysis published 2026-11-08] - [Browser and extension security: reduce the code surrounding every privileged session](https://www.web3sec.news/blog/browser-extension-security-runbook): Inventory permissions, separate sensitive browsing and prepare for malicious extensions or stolen browser sessions. [guide; analysis published 2026-11-08] - [GitHub and CI/CD security: keep untrusted changes away from deployment authority](https://www.web3sec.news/blog/github-ci-security-runbook): A review procedure for workflow triggers, action revisions, runner isolation and cloud deployment identity. [guide; analysis published 2026-11-08] - [Package publishing security: control who can ship and what consumers execute](https://www.web3sec.news/blog/package-publishing-security-runbook): A practical process for registry access, trusted publishing, provenance, dependency review and compromised releases. [guide; analysis published 2026-11-08] - [Cloud IAM and secrets: reduce standing access and prove revocation](https://www.web3sec.news/blog/cloud-iam-secrets-runbook): An operating model for human roles, workload identities, secret consumers and emergency cloud access. [guide; analysis published 2026-11-08] - [Production data and backups: prove you can recover without giving attackers a second copy](https://www.web3sec.news/blog/production-data-backup-runbook): A runbook for database access, backup isolation, retention, restoration and safe evidence during data incidents. [guide; analysis published 2026-11-08] - [API authorization: test the actor, object and action together](https://www.web3sec.news/blog/api-authorization-review-runbook): A practical review method for cross-tenant access, administrative functions, exports, background jobs and failure behavior. [guide; analysis published 2026-11-08] - [Multisig treasury operations: verify the transaction and every path around the threshold](https://www.web3sec.news/blog/multisig-treasury-security-runbook): An operating runbook for signer independence, privileged transactions, modules, monitoring and recovery rehearsal. [guide; analysis published 2026-11-08] - [Community and support security: protect announcements without becoming a recovery scam](https://www.web3sec.news/blog/community-support-security-runbook): Practical controls for moderators, bots, public links, support evidence and incident communications. [guide; analysis published 2026-11-08] - [Incident evidence: preserve what happened without spreading the breach](https://www.web3sec.news/blog/incident-evidence-runbook): A procedure for first-response records, log exports, file provenance, access restrictions and investigative handoff. [guide; analysis published 2026-11-08] - [Incident tabletop: rehearse decisions while the systems are still healthy](https://www.web3sec.news/blog/incident-tabletop-runbook): A repeatable exercise format for founders, engineering, security, support and the people authorized to stop operations. [guide; analysis published 2026-11-08] - [Security monitoring: connect important events to an owner who can act](https://www.web3sec.news/blog/security-monitoring-runbook): Build a small, testable detection system around identity changes, deployments, data access and treasury authority. [guide; analysis published 2026-11-08] - [Frontend security: review every script that can change what a user sees](https://www.web3sec.news/blog/frontend-third-party-script-security): A practical review of browser code, tag managers and content security policies, including the interactions a successful build can still break. [guide; analysis published 2026-11-08] - [Write a security finding someone else can verify](https://www.web3sec.news/blog/security-writing-evidence-review): A review procedure for findings, incident narratives and technical articles: keep evidence, inference, recommendations and uncertainty separate. [guide; analysis published 2026-11-08] - [Run unfamiliar research repositories without lending them your laptop’s credentials](https://www.web3sec.news/blog/untrusted-repository-research-sandbox): Inspect before executing, isolate dependencies and tests, constrain network and filesystem access, and verify disposal with dummy secrets. [guide; analysis published 2026-11-08] - [Scope a mobile security test: the app, the device and the API are separate boundaries](https://www.web3sec.news/blog/scope-mobile-app-security-test): Prepare Android and iOS builds, device coverage, representative accounts and acceptance criteria so a mobile assessment answers a real release question. [guide; analysis published 2026-11-08] - [Mobile deep links: verify the destination without trusting the requested action](https://www.web3sec.news/blog/mobile-deep-links-session-boundaries): Review Android App Links, iOS universal links, custom schemes and exported entry points against account state, input validation and server authorization. [guide; analysis published 2026-11-08] ## Incident studies - [Nomad: when an initialized default becomes permission to process a message](https://www.web3sec.news/blog/nomad-initialization-trust): A bridge authentication failure illustrates why deployment state, default values and upgrade behavior need a single security model. [postmortem; analysis published 2026-11-08; incident 2022-08-01] - [Euler V1: every path that changes collateral belongs in the solvency model](https://www.web3sec.news/blog/euler-donation-accounting): The March 2023 exploit connects an omitted account-health check to liquidation incentives and multi-step accounting invariants. [postmortem; analysis published 2026-11-08; incident 2023-03-13] - [Ledger Connect Kit: offboarding and runtime dependencies share a security boundary](https://www.web3sec.news/blog/ledger-connect-kit-publishing): A package-publishing compromise reached applications through dynamically loaded code without changing their own repositories. [postmortem; analysis published 2026-11-08; incident 2023-12-14] - [LastPass 2022: backup access can be the production-data incident](https://www.web3sec.news/blog/lastpass-vault-backups): The disclosed chain from development information to privileged endpoint compromise and stolen backups challenges narrow production boundaries. [postmortem; analysis published 2026-11-08; advisory 2022-12-22] - [Okta support system 2023: diagnostic files can carry live account authority](https://www.web3sec.news/blog/okta-support-session-tokens): A support-system compromise links browser synchronization, service-account access and sensitive HAR files to session hijacking. [postmortem; analysis published 2026-11-08; incident 2023-09-28] - [Bybit and Safe: the signing interface is part of your security boundary](https://www.web3sec.news/blog/bybit-safe-signing-boundary): The February 2025 theft shows why multiple signatures cannot compensate for a shared, compromised view of a transaction. [postmortem; analysis published 2026-11-08; incident 2025-02-21] - [Radiant Capital: a document request can become a treasury incident](https://www.web3sec.news/blog/radiant-capital-developer-opsec): A reported malware delivery chain connects developer communications, workstation trust and multisig operations. [postmortem; analysis published 2026-11-08; incident 2024-10-16] - [WazirX: make custody responsibilities testable before an incident](https://www.web3sec.news/blog/wazirx-custody-responsibility): The July 2024 wallet theft illustrates the operational consequences of unclear verification and response responsibilities. [postmortem; analysis published 2026-11-08; incident 2024-07-18] - [Snowflake customer intrusions: old credentials can remain a live attack path](https://www.web3sec.news/blog/snowflake-stolen-credentials): Mandiant’s 2024 investigation connects infostealer exposure, missing MFA and cloud data theft. [postmortem; analysis published 2026-11-08; incident 2024-06-10] - [Cloudflare’s Thanksgiving incident: a rotation checklist needs closure evidence](https://www.web3sec.news/blog/cloudflare-token-rotation): A 2024 disclosure about a 2023 intrusion shows how residual credentials can outlive an initial response. [postmortem; analysis published 2026-11-08; incident 2023-11-23] - [tj-actions compromise: CI logs can become a secret-exfiltration channel](https://www.web3sec.news/blog/tj-actions-ci-secret-exposure): CVE-2025-30066 demonstrates why action versions, workflow permissions and build logs belong in one threat model. [postmortem; analysis published 2026-11-08; incident 2025-03-14] - [XZ backdoor: review the artifact you deploy, not only the repository](https://www.web3sec.news/blog/xz-build-artifact-trust): CVE-2024-3094 illustrates the gap between source review, release packaging and the executable shipped downstream. [postmortem; analysis published 2026-11-08; incident 2024-03-29] ## Research notes - [Next.js October and November 2026 advisories: patch the deployed artifact](https://www.web3sec.news/blog/nextjs-october-november-2026-security-advisories): Recent framework advisories connect upstream libraries, image processing, caching and cookie handling to the software actually running in production. [analysis; analysis published 2026-11-08; advisory 2026-11-03] - [AI agents: keep untrusted content away from authorization decisions](https://www.web3sec.news/blog/ai-agents-prompt-injection-boundaries): A practical control model for assistants that read the web, inspect repositories and act through connected tools. [analysis; analysis published 2026-11-08] ## Optional text exports - [Full text collection](https://www.web3sec.news/llms-full.txt): Article and service text with sources, dates and canonical page URLs. - [Content metadata index](https://www.web3sec.news/content-index.json): A versioned metadata manifest; this is a site-specific format, not a universal protocol. - [Nomad: when an initialized default becomes permission to process a message](https://www.web3sec.news/articles/nomad-initialization-trust.md) - [Euler V1: every path that changes collateral belongs in the solvency model](https://www.web3sec.news/articles/euler-donation-accounting.md) - [Ledger Connect Kit: offboarding and runtime dependencies share a security boundary](https://www.web3sec.news/articles/ledger-connect-kit-publishing.md) - [LastPass 2022: backup access can be the production-data incident](https://www.web3sec.news/articles/lastpass-vault-backups.md) - [Okta support system 2023: diagnostic files can carry live account authority](https://www.web3sec.news/articles/okta-support-session-tokens.md) - [Next.js October and November 2026 advisories: patch the deployed artifact](https://www.web3sec.news/articles/nextjs-october-november-2026-security-advisories.md) - [Bybit and Safe: the signing interface is part of your security boundary](https://www.web3sec.news/articles/bybit-safe-signing-boundary.md) - [Radiant Capital: a document request can become a treasury incident](https://www.web3sec.news/articles/radiant-capital-developer-opsec.md) - [WazirX: make custody responsibilities testable before an incident](https://www.web3sec.news/articles/wazirx-custody-responsibility.md) - [Snowflake customer intrusions: old credentials can remain a live attack path](https://www.web3sec.news/articles/snowflake-stolen-credentials.md) - [Cloudflare’s Thanksgiving incident: a rotation checklist needs closure evidence](https://www.web3sec.news/articles/cloudflare-token-rotation.md) - [tj-actions compromise: CI logs can become a secret-exfiltration channel](https://www.web3sec.news/articles/tj-actions-ci-secret-exposure.md) - [XZ backdoor: review the artifact you deploy, not only the repository](https://www.web3sec.news/articles/xz-build-artifact-trust.md) - [AI agents: keep untrusted content away from authorization decisions](https://www.web3sec.news/articles/ai-agents-prompt-injection-boundaries.md) - [Identity security: phishing-resistant login, usable recovery and verified offboarding](https://www.web3sec.news/articles/identity-mfa-recovery-runbook.md) - [Email security: protect the recovery inbox and control who can send as your domain](https://www.web3sec.news/articles/email-account-security-runbook.md) - [Domain and DNS security: preserve control of the address your users trust](https://www.web3sec.news/articles/domain-dns-security-runbook.md) - [Developer device security: isolate experiments from production authority](https://www.web3sec.news/articles/developer-device-security-runbook.md) - [Browser and extension security: reduce the code surrounding every privileged session](https://www.web3sec.news/articles/browser-extension-security-runbook.md) - [GitHub and CI/CD security: keep untrusted changes away from deployment authority](https://www.web3sec.news/articles/github-ci-security-runbook.md) - [Package publishing security: control who can ship and what consumers execute](https://www.web3sec.news/articles/package-publishing-security-runbook.md) - [Cloud IAM and secrets: reduce standing access and prove revocation](https://www.web3sec.news/articles/cloud-iam-secrets-runbook.md) - [Production data and backups: prove you can recover without giving attackers a second copy](https://www.web3sec.news/articles/production-data-backup-runbook.md) - [API authorization: test the actor, object and action together](https://www.web3sec.news/articles/api-authorization-review-runbook.md) - [Multisig treasury operations: verify the transaction and every path around the threshold](https://www.web3sec.news/articles/multisig-treasury-security-runbook.md) - [Community and support security: protect announcements without becoming a recovery scam](https://www.web3sec.news/articles/community-support-security-runbook.md) - [AI agent and MCP security: constrain the tool, not just the prompt](https://www.web3sec.news/articles/ai-agent-mcp-security-runbook.md) - [Incident evidence: preserve what happened without spreading the breach](https://www.web3sec.news/articles/incident-evidence-runbook.md) - [Incident tabletop: rehearse decisions while the systems are still healthy](https://www.web3sec.news/articles/incident-tabletop-runbook.md) - [Security monitoring: connect important events to an owner who can act](https://www.web3sec.news/articles/security-monitoring-runbook.md) - [Frontend security: review every script that can change what a user sees](https://www.web3sec.news/articles/frontend-third-party-script-security.md) - [Write a security finding someone else can verify](https://www.web3sec.news/articles/security-writing-evidence-review.md) - [Connect an AI assistant to email and calendars without delegating your entire identity](https://www.web3sec.news/articles/personal-ai-assistant-email-calendar-security.md) - [Financial AI agents: separate research, transaction preparation and execution authority](https://www.web3sec.news/articles/financial-ai-agent-permission-boundaries.md) - [Run unfamiliar research repositories without lending them your laptop’s credentials](https://www.web3sec.news/articles/untrusted-repository-research-sandbox.md) - [AI-assisted bug bounty work: keep scope, execution and evidence under researcher control](https://www.web3sec.news/articles/ai-assisted-bug-bounty-workflow.md) - [Before installing an MCP server: review the package, process and connected account separately](https://www.web3sec.news/articles/mcp-server-installation-review.md) - [Agent credentials: issue the smallest useful secret and prove you can revoke it](https://www.web3sec.news/articles/agent-secrets-scope-and-recovery.md) - [Browser agents and logged-in sessions: protect the account behind the tab](https://www.web3sec.news/articles/browser-agent-session-security.md) - [Agent memory and RAG: prevent retrieved text from becoming persistent policy](https://www.web3sec.news/articles/agent-memory-retrieval-poisoning.md) - [Review agent skills and plugins as instructions, dependencies and delegated authority](https://www.web3sec.news/articles/agent-skills-plugin-review.md) - [Unattended agents: bound schedules, retries, cumulative usage and stale approvals](https://www.web3sec.news/articles/unattended-agent-schedules-and-budgets.md) - [Agents, localhost and internal networks: enforce what the executor can actually reach](https://www.web3sec.news/articles/agent-localhost-network-boundaries.md) - [Scope a mobile security test: the app, the device and the API are separate boundaries](https://www.web3sec.news/articles/scope-mobile-app-security-test.md) - [Mobile deep links: verify the destination without trusting the requested action](https://www.web3sec.news/articles/mobile-deep-links-session-boundaries.md) - [Smart contract security](https://www.web3sec.news/service-notes/smart-contract-security.md) - [Web & API pentesting](https://www.web3sec.news/service-notes/web-api-pentesting.md) - [Incident readiness & response](https://www.web3sec.news/service-notes/incident-response.md) - [AI & supply-chain security](https://www.web3sec.news/service-notes/ai-supply-chain-security.md) - [Technical writing & research](https://www.web3sec.news/service-notes/technical-writing.md) - [Security strategy & technical GTM](https://www.web3sec.news/service-notes/security-advisory.md) - [Mobile app security testing](https://www.web3sec.news/service-notes/mobile-app-security.md)