# AtlaSent

> AtlaSent Consequential Action Protection: control consequential actions by people, software, and AI before they run, and verify the proof afterward.

- [AtlaSent: Consequential Action Protection](https://docs.atlasent.io/index.md): AtlaSent controls consequential actions by people, software, and AI before they happen, and preserves proof of what was decided and what happened afterward.
- [Quickstart](https://docs.atlasent.io/quickstart/overview.md): Choose a safe AtlaSent quickstart and learn the authorization, Permit, verification, execution, and evidence boundaries.
- [Developer / API quickstart](https://docs.atlasent.io/quickstart/developer-api.md): Run a safe AtlaSent evaluation, handle a Permit, verify it, and inspect evidence without implying execution occurred.
- [1. Create an API key](https://docs.atlasent.io/quickstart/api-key.md): Generate a scoped API key in the AtlaSent Console.
- [2. Your first protected action](https://docs.atlasent.io/quickstart/first-evaluation.md): Acquire a verified Permit and execute the action under it. Fail-closed by construction — no manual decision branching.
- [3. Verify a permit explicitly](https://docs.atlasent.io/quickstart/verify-permit.md): The lower-level verify call — used when a permit crossed a boundary and the receiving side wants to re-confirm before acting.
- [4. Export the audit bundle](https://docs.atlasent.io/quickstart/audit-export.md): Download a signed, offline-verifiable audit bundle.
- [Console User Guide](https://docs.atlasent.io/console/index.md): Learn how to navigate the signed-in AtlaSent Console, interpret its fields, and complete the core authorization workflow.
- [Console Screens and Fields](https://docs.atlasent.io/console/screens-and-fields.md): A screen-by-screen reference for the primary signed-in AtlaSent Console, including the meaning and safe use of its common fields.
- [The Deploy Gate Step](https://docs.atlasent.io/deploy-gate.md): Fail-closed authorization for every production deployment — whether triggered by a human, CI/CD pipeline, or AI agent.
- [Get Started with Production Change Protection](https://docs.atlasent.io/production-change-gate-assessment.md): The directive checklist for trying, connecting, proving, and activating AtlaSent Production Change Protection.
- [GMP controls: authorize consequential manufacturing actions](https://docs.atlasent.io/gmp.md): Explore how AtlaSent's synthetic GMP testbed models execution-time authorization for batch release, regulated record changes, and process controls, and request a scoped demonstration.
- [AI Action Protection](https://docs.atlasent.io/ai-agent-governance.md): An AI agent's tool access is not permission to act. AtlaSent gates an agent's consequential actions before they run and keeps proof of what was decided.
- [Protected Actions](https://docs.atlasent.io/protected-actions.md): A protected action is any operation whose effect is real, external, and hard to reverse. These are the actions that require authorization before they run.
- [Permit Verification](https://docs.atlasent.io/permit-verification.md): How single-use permit tokens prove every authorized action was explicitly allowed — and can be verified offline.
- [How to Require Authorization Before an AI Agent Executes an Action](https://docs.atlasent.io/authorization-for-ai-agents.md): Place a permit-gated execution boundary in front of consequential AI-agent tool actions so verification occurs before effect.
- [Architecture](https://docs.atlasent.io/architecture.md): How AtlaSent works — the request evaluation flow, system components, fail-closed design, and key invariants.
- [Core Concepts](https://docs.atlasent.io/concepts/overview.md): The controlled concepts behind AtlaSent: verified identity, organizational authority, execution-time authorization, permits, gates, execution, effect, and proof.
- [Verified Actor Identity](https://docs.atlasent.io/concepts/verified-actor-identity.md): How AtlaSent establishes who or what is acting without confusing identity with organizational authority.
- [What Is Execution-Time Authorization?](https://docs.atlasent.io/concepts/execution-time-authorization.md): Execution-time authorization determines whether a consequential computational action is organizationally authorized at the execution boundary, before its effect occurs.
- [What Is Organizational Authority?](https://docs.atlasent.io/concepts/organizational-authority.md): Organizational authority is the standing, scoped organizational right to cause a class of change. It is distinct from technical access, approval, and per-request authorization.
- [What Is a Consequential Computational Action?](https://docs.atlasent.io/concepts/consequential-computational-actions.md): A consequential computational action can produce a material organizational effect and may require explicit authority, policy conditions, evidence, or enforcement before execution.
- [Identity vs. Authority vs. Authorization vs. Approval vs. Access](https://docs.atlasent.io/concepts/authority-vs-authorization.md): Identity, authority, authorization, approval, and access answer different questions. Keeping them separate is central to AtlaSent's organizational-authority model.
- [Runtime flow](https://docs.atlasent.io/concepts/runtime-flow.md): How AtlaSent moves from a proposed consequential action through organizational authority, authorization, Permit verification, execution, and evidence.
- [Glossary](https://docs.atlasent.io/concepts/glossary.md): Controlled AtlaSent vocabulary for organizational authority, execution-time authorization, protected actions, permits, gates, and evidence.
- [Action class](https://docs.atlasent.io/concepts/action-class.md): The typed action an agent requests authorization for.
- [Policy](https://docs.atlasent.io/concepts/policy.md): Versioned rules that decide the outcome for each action type.
- [Permit](https://docs.atlasent.io/concepts/permit.md): A Permit is the bounded runtime artifact that embodies a positive authorization and is verified at the execution boundary before the governed effect.
- [Audit chain](https://docs.atlasent.io/concepts/audit-chain.md): Hash-linked, Ed25519-signed, append-only record of every decision.
- [Proof verification](https://docs.atlasent.io/concepts/proof-verification.md): How AtlaSent makes audit trails tamper-evident, and how to verify a proof bundle offline using only the bundle, a public key, and the algorithm on this page.
- [Deny codes](https://docs.atlasent.io/concepts/deny-codes.md): Stable machine codes the engine returns when a decision is not allow. One page to point at when an operator sees a deny_code in the audit trail.
- [Override](https://docs.atlasent.io/concepts/override.md): A break-glass path for when the policy says deny but a human says yes.
- [Approval](https://docs.atlasent.io/concepts/approval.md): Human-in-the-loop gate for hold and escalate decisions.
- [What Can AtlaSent Authorize?](https://docs.atlasent.io/concepts/what-can-atlasent-authorize.md): AtlaSent authorizes consequential actions before they happen. One authorization model governs deployments, business-system changes, clinical actions, identity changes, and AI-agent execution — every surface mapping back to a canonical action.
- [The Board Constituency Map](https://docs.atlasent.io/concepts/board-constituency-map.md): Boards ask how to create value from AI while protecting customers, employees, intellectual property, and shareholder interests. That question has a concrete answer: the inventory of consequential actions that cannot execute without authorization, grouped by who each one protects.
- [Protected Action Canon](https://docs.atlasent.io/reference/protected-action-canon.md): The canonical set of consequential actions AtlaSent can govern — each with its stable identifier, slug, and the compliance obligations it helps satisfy.
- [Concept-to-Action Map](https://docs.atlasent.io/reference/concept-to-action-map.md): How AtlaSent's authorization concepts relate to the canonical protected actions they govern.
- [SDKs](https://docs.atlasent.io/sdks/overview.md): Official AtlaSent client libraries for TypeScript and Python, with Go in preview.
- [TypeScript SDK](https://docs.atlasent.io/sdks/typescript.md): @atlasent/sdk — Node.js and edge-runtime client for AtlaSent.
- [Python SDK](https://docs.atlasent.io/sdks/python.md): atlasent — Python client for AtlaSent.
- [Go SDK](https://docs.atlasent.io/sdks/go.md): The AtlaSent Go SDK is in preview and not yet published.
- [atlasent-policy CLI](https://docs.atlasent.io/sdks/cli.md): Keep AtlaSent policies as files in a repo: validate and test them offline, plan changes on pull requests, and apply on merge.
- [Integrations](https://docs.atlasent.io/integrations/overview.md): Drop AtlaSent into LangChain, LlamaIndex, MCP, GitHub Actions, Slack, and your SIEM.
- [GitHub Action](https://docs.atlasent.io/integrations/github-action.md): Gate production deploys on an AtlaSent permit.
- [GitHub App](https://docs.atlasent.io/integrations/github-app.md): Install the AtlaSent GitHub App to get Check Run status checks on pull requests — not just webhook event delivery.
- [MCP server](https://docs.atlasent.io/integrations/mcp-server.md): Authorize-before-execute for any MCP-compatible AI agent: evaluate an action, get a Permit, verify it before the side effect runs.
- [LangChain](https://docs.atlasent.io/integrations/langchain.md): Wrap LangChain tools so every tool call is authorized by AtlaSent before it executes.
- [LlamaIndex](https://docs.atlasent.io/integrations/llamaindex.md): Tool wrappers that inject AtlaSent authorization into LlamaIndex agents, in Python and TypeScript.
- [SIEM export](https://docs.atlasent.io/integrations/siem-export.md): Stream the AtlaSent audit chain into Splunk, Sumo, Datadog, or any Syslog-compatible endpoint.
- [SIEM Export Console](https://docs.atlasent.io/integrations/siem-export-console.md): Configure Splunk, Elastic, or generic webhook destinations for AtlaSent audit events from the Console UI.
- [Slack Approval Bot](https://docs.atlasent.io/integrations/slack-approval-bot.md): Route AtlaSent hold and escalate decisions to a Slack thread for human approval using Block Kit interactive messages.
- [V1 pilot migration](https://docs.atlasent.io/guides/v1-pilot-migration.md): Wire and SDK changes for callers moving onto AtlaSent V1 from earlier builds
- [SSO Setup](https://docs.atlasent.io/guides/sso-setup.md): Configure and acceptance-test enterprise SAML SSO for an AtlaSent organization; OIDC remains limited/early access.
- [Webhook subscriptions](https://docs.atlasent.io/guides/webhooks.md): Create a webhook subscription, verify the HMAC-SHA256 signature on each delivery, and inspect or replay failed deliveries.
- [AtlaSent Enterprise](https://docs.atlasent.io/enterprise/index.md): AtlaSent Enterprise is the operating layer for Consequential Action Protection across teams and systems. This section documents its identity, export and evidence capabilities.
- [SCIM 2.0 Provisioning](https://docs.atlasent.io/enterprise/scim-admin.md): Configure and acceptance-test AtlaSent's enterprise SCIM 2.0 provisioning surface.
- [SIEM Export (Enterprise)](https://docs.atlasent.io/enterprise/siem-export.md): Schedule audit event exports to Splunk, Datadog, or Elastic from the AtlaSent Enterprise control center.
- [Governed close operations](https://docs.atlasent.io/runbooks/governed-close-operations.md): End-to-end runbook for AtlaSent gating accounting close: evaluate → permit → verify → execute → consume → audit, with worked examples for journal-entry posting, vendor master changes, period lock/unlock, approval escalation, and emergency override with tamper-evident evidence.
- [Compliance](https://docs.atlasent.io/compliance/overview.md): Which regulatory frameworks AtlaSent aligns with, and where to verify each claim.
- [21 CFR Part 11](https://docs.atlasent.io/compliance/21-cfr-part-11.md): FDA regulation on electronic records and electronic signatures — which section maps to which AtlaSent primitive.
- [EU Annex 11](https://docs.atlasent.io/compliance/eu-annex-11.md): EMA guideline for computerized systems in GMP — which clause maps to which AtlaSent primitive.
- [ALCOA+](https://docs.atlasent.io/compliance/alcoa-plus.md): Data integrity principles — how each maps to an AtlaSent primitive.
- [ICH E6 GCP](https://docs.atlasent.io/compliance/ich-e6-gcp.md): International GCP for clinical trials — which section maps to which AtlaSent primitive.
- [GxP change control](https://docs.atlasent.io/compliance/gxp-change-control.md): Execution-time authorization patterns for regulated system changes, their quality-system evidence, and what must be proven before a customer enforcement claim.
- [AtlaSent Trust Portal](https://docs.atlasent.io/trust/index.md): Security, compliance, legal, and operational evidence for AtlaSent customers and prospects — one URL to share with procurement.
- [Security overview](https://docs.atlasent.io/trust/security-overview.md): How AtlaSent protects customer data — encryption, access controls, audit-chain integrity, and fail-closed guarantees.
- [Compliance status](https://docs.atlasent.io/trust/compliance-status.md): Current certifications, frameworks, roadmap. Calendar view of when SOC 2, HIPAA BAA, and other gates land.
- [Sub-processors](https://docs.atlasent.io/trust/subprocessors.md): Every third party that processes Customer Personal Data on AtlaSent's behalf, with locations and cross-border transfer mechanisms. 30-day change notification.
- [Audit signing key](https://docs.atlasent.io/trust/audit-signing-key.md): AtlaSent's public Ed25519 audit-entry signing key. Pin it out-of-band and verify any exported audit chain or proof bundle offline — against a key you got from us, not from the bundle.
- [Contact](https://docs.atlasent.io/trust/contact.md): How to reach AtlaSent for support, security, privacy, procurement, legal, and general inquiries.
- [For Biotech CCOs](https://docs.atlasent.io/biotech/overview.md): A short, honest read for Chief Compliance Officers and Quality leads at biotech and pharma companies evaluating AtlaSent.
- [The thesis](https://docs.atlasent.io/biotech/thesis.md): Why execution-time authorization is the piece the market is missing.
- [API Reference](https://docs.atlasent.io/api-reference/introduction.md): The canonical AtlaSent REST API, rendered from the OpenAPI specification.
- [Authentication](https://docs.atlasent.io/api-reference/authentication.md): JWT bearer for user sessions, API keys for service-to-service.
- [Roles Reference](https://docs.atlasent.io/api-reference/roles.md): RBAC roles for the self-hosted AtlaSent control plane and how to assign them.
- [Rate limits](https://docs.atlasent.io/api-reference/rate-limits.md): Per-org quotas and how to handle 429 responses.
- [Errors](https://docs.atlasent.io/api-reference/errors.md): Error codes returned by the AtlaSent API and how to recover from each.
- [Pagination](https://docs.atlasent.io/api-reference/pagination.md): Cursor-based pagination, stable under inserts.
- [Versioning & Compatibility](https://docs.atlasent.io/api-reference/versioning.md): How AtlaSent versions its API, what stability guarantees apply to each endpoint category, what additive evolution looks like, and what to expect when things change.
- [HITL Escalations](https://docs.atlasent.io/api-reference/escalations.md): Human-in-the-loop escalation endpoints: create, list, inspect, and respond to escalations.
- [Changelog](https://docs.atlasent.io/changelog/overview.md): Release notes for the AtlaSent platform: API, Console, and SDKs.
- [API changelog](https://docs.atlasent.io/changelog/api.md): Release notes for atlasent-api, synced from its customer-facing changelog.
- [Console changelog](https://docs.atlasent.io/changelog/console.md): Release notes for atlasent-console, synced from its customer-facing changelog.
- [SDK changelogs](https://docs.atlasent.io/changelog/sdks.md): Release notes for TypeScript, Python, and Go SDKs.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.