# 2501.ai - [0.20 (176 pages)](https://docs.2501.ai/_llms/0-20.txt): Docs, Reference, Benchmark, Guides. - [0.19 (164 pages)](https://docs.2501.ai/_llms/0-19.txt): Docs, Reference, Benchmark, Guides. - [0.18 (161 pages)](https://docs.2501.ai/_llms/0-18.txt): Docs, Reference, Benchmark, Guides. - [0.17 (167 pages)](https://docs.2501.ai/_llms/0-17.txt): Docs, Reference, Benchmark, Guides. - [0.16 (143 pages)](https://docs.2501.ai/_llms/0-16.txt): Docs, Reference, Benchmark, Guides. - [0.15 (140 pages)](https://docs.2501.ai/_llms/0-15.txt): Docs, Reference, Benchmark, Guides. - [0.14 (126 pages)](https://docs.2501.ai/_llms/0-14.txt): Docs, Reference, Benchmark, Guides. ## 0.13 ### Docs #### Getting Started - [Welcome](https://docs.2501.ai/0.13/getting-started/welcome.md): AIOps with 2501 — an autonomous agent platform for IT operations - [Quickstart](https://docs.2501.ai/0.13/getting-started/quickstart.md): Install the CLI, deploy 2501, and run your first agent #### Understand 2501 - [Command Center](https://docs.2501.ai/0.13/understand/command-center.md): The web UI for managing your 2501 installation - [Engine & Agents](https://docs.2501.ai/0.13/understand/engine-agents.md): The backend that powers 2501 and the dual-LLM architecture behind every agent - [Agentic Flow](https://docs.2501.ai/0.13/understand/agentic-flow.md): Ticket → gateway → job → tasks → resolution, in one page - [Configuration Approaches](https://docs.2501.ai/0.13/understand/configuration.md): Three ways to configure 2501 — pick the one that fits how you work #### Concepts - [Agents](https://docs.2501.ai/0.13/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.13/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.13/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tickets](https://docs.2501.ai/0.13/core-concepts/tickets.md): Key data extracted from your ITSM, translated for 2501's pipeline - [Tasks](https://docs.2501.ai/0.13/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.13/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Job Schedules](https://docs.2501.ai/0.13/core-concepts/job-schedules.md): Recurring jobs that fan out into a fresh Job each tick - [Gateways](https://docs.2501.ai/0.13/core-concepts/gateways.md): Connect agents to external services ### Reference #### API - [API Overview](https://docs.2501.ai/0.13/api/overview.md): Drive 2501 as code over the versioned /api/v1 HTTP API ##### Specialties - [List specialties](https://docs.2501.ai/0.13/api-reference/specialties/list-specialties.md): org_id filters to that org plus the tenant-wide rows; omitting it lists the whole accessible scope. - [Create a specialty](https://docs.2501.ai/0.13/api-reference/specialties/create-a-specialty.md): key is derived from name. Omitting org_id creates a tenant-wide specialty and takes tenant-level access. - [Search specialties](https://docs.2501.ai/0.13/api-reference/specialties/search-specialties.md) - [Get a specialty](https://docs.2501.ai/0.13/api-reference/specialties/get-a-specialty.md) - [Update a specialty](https://docs.2501.ai/0.13/api-reference/specialties/update-a-specialty.md): Partial update. org_id names the scope the row moves TO, authorized like a create; writing a tenant-wide row takes tenant-level access. A caller that can see the row but not write its scope gets 403, not 404. - [Delete a specialty](https://docs.2501.ai/0.13/api-reference/specialties/delete-a-specialty.md): A specialty still carried by an agent is 409, not deleted. - [Count what references a specialty](https://docs.2501.ai/0.13/api-reference/specialties/count-what-references-a-specialty.md): A non-zero count means a delete would orphan dependents. ##### Blacklists - [List blacklisted commands](https://docs.2501.ai/0.13/api-reference/blacklists/list-blacklisted-commands.md): org_id filters to that org plus the tenant-wide rows; omitting it lists the whole accessible scope. - [Create a blacklisted command](https://docs.2501.ai/0.13/api-reference/blacklists/create-a-blacklisted-command.md): Omitting org_id creates a tenant-wide entry and takes tenant-level access. - [Search blacklisted commands](https://docs.2501.ai/0.13/api-reference/blacklists/search-blacklisted-commands.md): q matches the id, the pattern and the description. - [Get a blacklisted command](https://docs.2501.ai/0.13/api-reference/blacklists/get-a-blacklisted-command.md) - [Update a blacklisted command](https://docs.2501.ai/0.13/api-reference/blacklists/update-a-blacklisted-command.md): Partial update. org_id names the scope the row moves TO, authorized like a create; writing a tenant-wide row takes tenant-level access. A caller that can see the row but not write its scope gets 403, not 404. - [Delete a blacklisted command](https://docs.2501.ai/0.13/api-reference/blacklists/delete-a-blacklisted-command.md) ##### Hosts - [List hosts](https://docs.2501.ai/0.13/api-reference/hosts/list-hosts.md) - [Create a host](https://docs.2501.ai/0.13/api-reference/hosts/create-a-host.md): Success is 200, not 201. Host names are not enforced unique. - [Search hosts](https://docs.2501.ai/0.13/api-reference/hosts/search-hosts.md) - [Bulk import hosts](https://docs.2501.ai/0.13/api-reference/hosts/bulk-import-hosts.md): Body is a CSV file (text/csv) or a JSON envelope. Returns 200 with a per-row verdict even when every row failed. - [Bulk export hosts](https://docs.2501.ai/0.13/api-reference/hosts/bulk-export-hosts.md): Accept: text/csv gets the file; anything else gets the same rows as a JSON array (what batch takes as rows). - [Get a host](https://docs.2501.ai/0.13/api-reference/hosts/get-a-host.md): Adds an `agents` array of the host usable agents. - [Update a host](https://docs.2501.ai/0.13/api-reference/hosts/update-a-host.md): Partial update - only the fields in the body change. - [Delete a host](https://docs.2501.ai/0.13/api-reference/hosts/delete-a-host.md): Archives the host agents and their tasks, then deletes the host, in one transaction. - [List a host agents](https://docs.2501.ai/0.13/api-reference/hosts/list-a-host-agents.md) ##### Agents - [List agents](https://docs.2501.ai/0.13/api-reference/agents/list-agents.md) - [Create an agent](https://docs.2501.ai/0.13/api-reference/agents/create-an-agent.md) - [Search agents](https://docs.2501.ai/0.13/api-reference/agents/search-agents.md) - [Bulk import agents](https://docs.2501.ai/0.13/api-reference/agents/bulk-import-agents.md): Import hosts first: an agent row points at a host by id. - [Bulk export agents](https://docs.2501.ai/0.13/api-reference/agents/bulk-export-agents.md) - [Get an agent](https://docs.2501.ai/0.13/api-reference/agents/get-an-agent.md) - [Update an agent](https://docs.2501.ai/0.13/api-reference/agents/update-an-agent.md): Partial update. host_id and plugin_ids are not accepted here and are dropped silently. - [Archive an agent](https://docs.2501.ai/0.13/api-reference/agents/archive-an-agent.md): This is the delete. An archived agent stops appearing in reads. - [Archive idle tasks](https://docs.2501.ai/0.13/api-reference/agents/archive-idle-tasks.md): Archives the agent idle tasks only; nothing in flight is interrupted. - [List an agent tasks](https://docs.2501.ai/0.13/api-reference/agents/list-an-agent-tasks.md): Live tasks by default; archived=true for the archived list. status filters the archived list only. - [Test the connection](https://docs.2501.ai/0.13/api-reference/agents/test-the-connection.md): The engine tries a real connection. A failed login is 200 {success:false}; an unreachable engine is a 5xx ENGINE_ERROR. - [List an agent plugins](https://docs.2501.ai/0.13/api-reference/agents/list-an-agent-plugins.md) - [Attach a plugin](https://docs.2501.ai/0.13/api-reference/agents/attach-a-plugin.md): Returns the agent, not the plugin list. Re-read /plugins for the new set. - [Detach a plugin](https://docs.2501.ai/0.13/api-reference/agents/detach-a-plugin.md): Returns the agent, not the plugin list. #### CLI - [2501 CLI](https://docs.2501.ai/0.13/cli/overview.md): One binary, six command groups — manage the platform from your terminal or CI - [login & status](https://docs.2501.ai/0.13/cli/login-status.md): Sign in to your Command Center, confirm who you are - [resources](https://docs.2501.ai/0.13/cli/resources.md): Manage platform resources as version-controlled MDX files - [infra](https://docs.2501.ai/0.13/cli/infra.md): Deploy and operate 2501 on Docker Swarm or Kubernetes - [runner](https://docs.2501.ai/0.13/cli/runner.md): Drive Benchmark scenarios from the CLI - [chat](https://docs.2501.ai/0.13/cli/chat.md): Talk to the AI Assistant from your terminal #### Deploy - [Overview](https://docs.2501.ai/0.13/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.13/deployment/init-command.md): Initialize your 2501 deployment workspace - [Configuration](https://docs.2501.ai/0.13/deployment/configuration.md): How 2501 deployment configuration works - [Docker Swarm](https://docs.2501.ai/0.13/deployment/docker-swarm.md): Production deployment with Docker Swarm - [Kubernetes](https://docs.2501.ai/0.13/deployment/kubernetes.md): Run the 2501 engine on any Kubernetes cluster with the Kubernetes pod executor backend - [LLM Providers](https://docs.2501.ai/0.13/deployment/llm-providers.md): Seed LLM providers at deploy time and manage them in Command Center - [Users & Organizations](https://docs.2501.ai/0.13/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.13/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 - [Feature Toggles](https://docs.2501.ai/0.13/deployment/feature-toggles.md): Toggle experimental and optional behaviors on or off #### Configure - [Clusters](https://docs.2501.ai/0.13/configure/clusters.md): Run task executors inside your own Kubernetes cluster to reach isolated hosts - [Knowledge Base](https://docs.2501.ai/0.13/configure/knowledge.md): Upload documents and let the engine extract reusable knowledge for your agents - [Import and Export](https://docs.2501.ai/0.13/configure/import-export.md): Bulk import and export hosts and agents as CSV or JSON - [Users](https://docs.2501.ai/0.13/configure/users.md): Manage users and permissions from the Command Center - [Licensing & Usage](https://docs.2501.ai/0.13/configure/licensing.md): Apply your 2501 license, understand entitlement caps, and track usage from Command Center ##### Agent Behavior - [Specialties](https://docs.2501.ai/0.13/configure/specialties.md): Configure your agent for specific task domains - [Operational Rules](https://docs.2501.ai/0.13/configure/operational-rules.md): Tag-matched guardrails that apply to the right tasks, not every agent - [Blacklist](https://docs.2501.ai/0.13/configure/blacklist.md): Prevent agents from executing specific commands ##### Access & Secrets - [Credentials](https://docs.2501.ai/0.13/configure/credentials.md): Enable agents to access your services - [API Keys](https://docs.2501.ai/0.13/configure/api-keys.md): Generate scoped keys to call the 2501 API programmatically - [Plugins](https://docs.2501.ai/0.13/configure/plugins.md): Connect AI Agents to MCPs and plugins ##### Integration - [Webhooks](https://docs.2501.ai/0.13/configure/webhooks.md): Set up real-time webhook notifications from ServiceNow - [Configuration as Code](https://docs.2501.ai/0.13/configure/configuration-as-code.md): Manage platform resources as version-controlled MDX with the 2501 CLI - [Network Discovery](https://docs.2501.ai/0.13/configure/discovery.md): Scan your network to build host inventory automatically ###### Resource Reference - [Agent](https://docs.2501.ai/0.13/configure/resources/agents.md): MDX frontmatter reference for agent resources - [Host](https://docs.2501.ai/0.13/configure/resources/hosts.md): MDX frontmatter reference for host resources - [Specialty](https://docs.2501.ai/0.13/configure/resources/specialties.md): MDX frontmatter reference for specialty resources - [Operational Rule](https://docs.2501.ai/0.13/configure/resources/operational-rules.md): MDX frontmatter reference for operational rule resources - [Credential](https://docs.2501.ai/0.13/configure/resources/credentials.md): MDX frontmatter reference for credential resources - [Gateway](https://docs.2501.ai/0.13/configure/resources/gateways.md): MDX frontmatter reference for gateway resources - [Blacklist](https://docs.2501.ai/0.13/configure/resources/blacklist.md): MDX frontmatter reference for blacklist entries ##### LLM - [Providers](https://docs.2501.ai/0.13/configure/providers.md): Register the LLM endpoints the engine can call - [Models](https://docs.2501.ai/0.13/configure/models.md): Catalog the LLMs your agents can pick as engines ### Benchmark #### Get Started - [Why Benchmark](https://docs.2501.ai/0.13/benchmark/overview.md): Validate agent behavior on realistic scenarios before exposing it to production - [Quickstart](https://docs.2501.ai/0.13/benchmark/quickstart.md): Run your first benchmark scenario end to end #### Authoring - [Scenario Structure](https://docs.2501.ai/0.13/benchmark/scenario.md): Scenario structure, keys, tags, and writing effective descriptions - [Hosts](https://docs.2501.ai/0.13/benchmark/hosts.md): Target host configuration for host mode and VM mode - [Agents](https://docs.2501.ai/0.13/benchmark/agents.md): Agent configuration for host mode and VM mode - [Playbooks](https://docs.2501.ai/0.13/benchmark/playbooks.md): Ansible playbooks for prepare, restore, and validate phases - [Validation](https://docs.2501.ai/0.13/benchmark/validation.md): Validation rules, validators, and scoring model - [Examples](https://docs.2501.ai/0.13/benchmark/examples.md): Complete worked scenarios ready to copy and adapt #### Running - [start](https://docs.2501.ai/0.13/benchmark/start.md): CLI reference for 2501 runner start - [validate](https://docs.2501.ai/0.13/benchmark/validate.md): Lint scenario definitions or re-run validation rules against existing runs - [flush](https://docs.2501.ai/0.13/benchmark/flush.md): Delete scenario run data from the database #### VM Sandbox - [Overview](https://docs.2501.ai/0.13/benchmark/sandbox.md): Isolated, ephemeral VMs for scenario runs and agent testing - [Templates](https://docs.2501.ai/0.13/benchmark/sandbox-templates.md): VM base images available for incus and lima modes - [From a Scenario](https://docs.2501.ai/0.13/benchmark/sandbox-scenario.md): Step into a broken scenario environment to inspect state and iterate on playbooks - [Standalone VMs](https://docs.2501.ai/0.13/benchmark/sandbox-targets.md): Provision standalone VMs for ad-hoc agent testing outside of scenario runs ### Guides #### Prompting - [Prompting](https://docs.2501.ai/0.13/configure/prompting.md): Learn the basics of prompting AI Agents - [Talking to an Agent](https://docs.2501.ai/0.13/prompting/talking-to-agents.md): How to write task prompts that agents actually act on correctly - [Prompting for a Specialty](https://docs.2501.ai/0.13/prompting/specialty.md): Shape how an agent thinks about a domain — vocabulary, priorities, and approach - [Prompting an Operational Rule](https://docs.2501.ai/0.13/prompting/operational-rule.md): Tell the agent what is true about *your* environment - [Prompting a Gateway](https://docs.2501.ai/0.13/prompting/gateway.md): Route and sequence inbound tickets, and govern what 2501 writes back #### Best Practices - [Scoping Agents](https://docs.2501.ai/0.13/best-practices/scoping-agents.md): Why focused agents beat general-purpose ones — and how to scope them - [Working with Knowledge](https://docs.2501.ai/0.13/best-practices/knowledge.md): Feed your existing documentation into 2501 so agents inherit it automatically - [Dynamic Context Injection](https://docs.2501.ai/0.13/best-practices/dynamic-context.md): How 2501 chooses what to put in an agent's context — and what stays out - [MCP & Plugins](https://docs.2501.ai/0.13/best-practices/mcp-plugins.md): When to give agents extra tools — and when not to - [Testing a Behavior](https://docs.2501.ai/0.13/best-practices/testing.md): Verify a new specialty, rule, or MCP before rolling it out to production #### Risk & Safety - [Handling Permissions](https://docs.2501.ai/0.13/risk/permissions.md): Limit agents at the credential and provisioning layer, not just the prompt - [Blacklists in Practice](https://docs.2501.ai/0.13/risk/blacklists.md): The programmatic kill-switch — always on, no matter what - [Read-Only Agents](https://docs.2501.ai/0.13/risk/read-only-agents.md): Use the investigate-only ceiling for critical systems and high-risk new use cases - [Benchmarking for Risk](https://docs.2501.ai/0.13/risk/benchmarking.md): Validate new use cases in a sandbox before exposing them to production #### FAQ - [FAQ](https://docs.2501.ai/0.13/faq.md): Quick answers to common questions about 2501 ## 0.12 ### Docs #### Getting Started - [Welcome](https://docs.2501.ai/0.12/getting-started/welcome.md): AIOps with 2501 — an autonomous agent platform for IT operations - [Quickstart](https://docs.2501.ai/0.12/getting-started/quickstart.md): Install the CLI, deploy 2501, and run your first agent #### Understand 2501 - [Command Center](https://docs.2501.ai/0.12/understand/command-center.md): The web UI for managing your 2501 installation - [Engine & Agents](https://docs.2501.ai/0.12/understand/engine-agents.md): The backend that powers 2501 and the dual-LLM architecture behind every agent - [Agentic Flow](https://docs.2501.ai/0.12/understand/agentic-flow.md): Ticket → gateway → job → tasks → resolution, in one page - [Configuration Approaches](https://docs.2501.ai/0.12/understand/configuration.md): Three ways to configure 2501 — pick the one that fits how you work #### Concepts - [Agents](https://docs.2501.ai/0.12/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.12/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.12/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tickets](https://docs.2501.ai/0.12/core-concepts/tickets.md): Key data extracted from your ITSM, translated for 2501's pipeline - [Tasks](https://docs.2501.ai/0.12/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.12/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Job Schedules](https://docs.2501.ai/0.12/core-concepts/job-schedules.md): Recurring jobs that fan out into a fresh Job each tick - [Gateways](https://docs.2501.ai/0.12/core-concepts/gateways.md): Connect agents to external services ### Reference #### API - [API Overview](https://docs.2501.ai/0.12/api/overview.md): Drive 2501 as code over the versioned /api/v1 HTTP API ##### Hosts - [List hosts](https://docs.2501.ai/0.12/api-reference/hosts/list-hosts.md) - [Create a host](https://docs.2501.ai/0.12/api-reference/hosts/create-a-host.md): Success is 200, not 201. Host names are not enforced unique. - [Search hosts](https://docs.2501.ai/0.12/api-reference/hosts/search-hosts.md) - [Bulk import hosts](https://docs.2501.ai/0.12/api-reference/hosts/bulk-import-hosts.md): Body is a CSV file (text/csv) or a JSON envelope. Returns 200 with a per-row verdict even when every row failed. - [Bulk export hosts](https://docs.2501.ai/0.12/api-reference/hosts/bulk-export-hosts.md): Accept: text/csv gets the file; anything else gets the same rows as a JSON array (what batch takes as rows). - [Get a host](https://docs.2501.ai/0.12/api-reference/hosts/get-a-host.md): Adds an `agents` array of the host usable agents. - [Update a host](https://docs.2501.ai/0.12/api-reference/hosts/update-a-host.md): Partial update - only the fields in the body change. - [Delete a host](https://docs.2501.ai/0.12/api-reference/hosts/delete-a-host.md): Archives the host agents and their tasks, then deletes the host, in one transaction. - [List a host agents](https://docs.2501.ai/0.12/api-reference/hosts/list-a-host-agents.md) ##### Agents - [List agents](https://docs.2501.ai/0.12/api-reference/agents/list-agents.md) - [Create an agent](https://docs.2501.ai/0.12/api-reference/agents/create-an-agent.md) - [Search agents](https://docs.2501.ai/0.12/api-reference/agents/search-agents.md) - [Bulk import agents](https://docs.2501.ai/0.12/api-reference/agents/bulk-import-agents.md): Import hosts first: an agent row points at a host by id. - [Bulk export agents](https://docs.2501.ai/0.12/api-reference/agents/bulk-export-agents.md) - [Get an agent](https://docs.2501.ai/0.12/api-reference/agents/get-an-agent.md) - [Update an agent](https://docs.2501.ai/0.12/api-reference/agents/update-an-agent.md): Partial update. host_id and plugin_ids are not accepted here and are dropped silently. - [Archive an agent](https://docs.2501.ai/0.12/api-reference/agents/archive-an-agent.md): This is the delete. An archived agent stops appearing in reads. - [Archive idle tasks](https://docs.2501.ai/0.12/api-reference/agents/archive-idle-tasks.md): Archives the agent idle tasks only; nothing in flight is interrupted. - [List an agent tasks](https://docs.2501.ai/0.12/api-reference/agents/list-an-agent-tasks.md): Live tasks by default; archived=true for the archived list. status filters the archived list only. - [Test the connection](https://docs.2501.ai/0.12/api-reference/agents/test-the-connection.md): The engine tries a real connection. A failed login is 200 {success:false}; an unreachable engine is a 5xx ENGINE_ERROR. - [List an agent plugins](https://docs.2501.ai/0.12/api-reference/agents/list-an-agent-plugins.md) - [Attach a plugin](https://docs.2501.ai/0.12/api-reference/agents/attach-a-plugin.md): Returns the agent, not the plugin list. Re-read /plugins for the new set. - [Detach a plugin](https://docs.2501.ai/0.12/api-reference/agents/detach-a-plugin.md): Returns the agent, not the plugin list. #### CLI - [2501 CLI](https://docs.2501.ai/0.12/cli/overview.md): One binary, six command groups — manage the platform from your terminal or CI - [login & status](https://docs.2501.ai/0.12/cli/login-status.md): Sign in to your Command Center, confirm who you are - [resources](https://docs.2501.ai/0.12/cli/resources.md): Manage platform resources as version-controlled MDX files - [infra](https://docs.2501.ai/0.12/cli/infra.md): Deploy and operate 2501 on Docker Swarm or Kubernetes - [runner](https://docs.2501.ai/0.12/cli/runner.md): Drive Benchmark scenarios from the CLI - [chat](https://docs.2501.ai/0.12/cli/chat.md): Talk to the AI Assistant from your terminal #### Deploy - [Overview](https://docs.2501.ai/0.12/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.12/deployment/init-command.md): Initialize your 2501 deployment workspace - [Configuration](https://docs.2501.ai/0.12/deployment/configuration.md): How 2501 deployment configuration works - [Docker Swarm](https://docs.2501.ai/0.12/deployment/docker-swarm.md): Production deployment with Docker Swarm - [Kubernetes](https://docs.2501.ai/0.12/deployment/kubernetes.md): Run the 2501 engine on any Kubernetes cluster with the Kubernetes pod executor backend - [LLM Providers](https://docs.2501.ai/0.12/deployment/llm-providers.md): Seed LLM providers at deploy time and manage them in Command Center - [Users & Organizations](https://docs.2501.ai/0.12/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.12/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 - [Feature Toggles](https://docs.2501.ai/0.12/deployment/feature-toggles.md): Toggle experimental and optional behaviors on or off #### Configure - [Knowledge Base](https://docs.2501.ai/0.12/configure/knowledge.md): Upload documents and let the engine extract reusable knowledge for your agents - [Import and Export](https://docs.2501.ai/0.12/configure/import-export.md): Bulk import and export hosts and agents as CSV or JSON - [Users](https://docs.2501.ai/0.12/configure/users.md): Manage users and permissions from the Command Center - [Licensing & Usage](https://docs.2501.ai/0.12/configure/licensing.md): Apply your 2501 license, understand entitlement caps, and track usage from Command Center ##### Agent Behavior - [Specialties](https://docs.2501.ai/0.12/configure/specialties.md): Configure your agent for specific task domains - [Operational Rules](https://docs.2501.ai/0.12/configure/operational-rules.md): Tag-matched guardrails that apply to the right tasks, not every agent - [Blacklist](https://docs.2501.ai/0.12/configure/blacklist.md): Prevent agents from executing specific commands ##### Access & Secrets - [Credentials](https://docs.2501.ai/0.12/configure/credentials.md): Enable agents to access your services - [API Keys](https://docs.2501.ai/0.12/configure/api-keys.md): Generate scoped keys to call the 2501 API programmatically - [Plugins](https://docs.2501.ai/0.12/configure/plugins.md): Connect AI Agents to MCPs and plugins ##### Integration - [Webhooks](https://docs.2501.ai/0.12/configure/webhooks.md): Set up real-time webhook notifications from ServiceNow - [Configuration as Code](https://docs.2501.ai/0.12/configure/configuration-as-code.md): Manage platform resources as version-controlled MDX with the 2501 CLI - [Network Discovery](https://docs.2501.ai/0.12/configure/discovery.md): Scan your network to build host inventory automatically ###### Resource Reference - [Agent](https://docs.2501.ai/0.12/configure/resources/agents.md): MDX frontmatter reference for agent resources - [Host](https://docs.2501.ai/0.12/configure/resources/hosts.md): MDX frontmatter reference for host resources - [Specialty](https://docs.2501.ai/0.12/configure/resources/specialties.md): MDX frontmatter reference for specialty resources - [Operational Rule](https://docs.2501.ai/0.12/configure/resources/operational-rules.md): MDX frontmatter reference for operational rule resources - [Credential](https://docs.2501.ai/0.12/configure/resources/credentials.md): MDX frontmatter reference for credential resources - [Gateway](https://docs.2501.ai/0.12/configure/resources/gateways.md): MDX frontmatter reference for gateway resources - [Blacklist](https://docs.2501.ai/0.12/configure/resources/blacklist.md): MDX frontmatter reference for blacklist entries ##### LLM - [Providers](https://docs.2501.ai/0.12/configure/providers.md): Register the LLM endpoints the engine can call - [Models](https://docs.2501.ai/0.12/configure/models.md): Catalog the LLMs your agents can pick as engines ### Benchmark #### Get Started - [Why Benchmark](https://docs.2501.ai/0.12/benchmark/overview.md): Validate agent behavior on realistic scenarios before exposing it to production - [Quickstart](https://docs.2501.ai/0.12/benchmark/quickstart.md): Run your first benchmark scenario end to end #### Authoring - [Scenario Structure](https://docs.2501.ai/0.12/benchmark/scenario.md): Scenario structure, keys, tags, and writing effective descriptions - [Hosts](https://docs.2501.ai/0.12/benchmark/hosts.md): Target host configuration for host mode and VM mode - [Agents](https://docs.2501.ai/0.12/benchmark/agents.md): Agent configuration for host mode and VM mode - [Playbooks](https://docs.2501.ai/0.12/benchmark/playbooks.md): Ansible playbooks for prepare, restore, and validate phases - [Validation](https://docs.2501.ai/0.12/benchmark/validation.md): Validation rules, validators, and scoring model - [Examples](https://docs.2501.ai/0.12/benchmark/examples.md): Complete worked scenarios ready to copy and adapt #### Running - [start](https://docs.2501.ai/0.12/benchmark/start.md): CLI reference for 2501 runner start - [validate](https://docs.2501.ai/0.12/benchmark/validate.md): Lint scenario definitions or re-run validation rules against existing runs - [flush](https://docs.2501.ai/0.12/benchmark/flush.md): Delete scenario run data from the database #### VM Sandbox - [Overview](https://docs.2501.ai/0.12/benchmark/sandbox.md): Isolated, ephemeral VMs for scenario runs and agent testing - [Templates](https://docs.2501.ai/0.12/benchmark/sandbox-templates.md): VM base images available for incus and lima modes - [From a Scenario](https://docs.2501.ai/0.12/benchmark/sandbox-scenario.md): Step into a broken scenario environment to inspect state and iterate on playbooks - [Standalone VMs](https://docs.2501.ai/0.12/benchmark/sandbox-targets.md): Provision standalone VMs for ad-hoc agent testing outside of scenario runs ### Guides #### Prompting - [Prompting](https://docs.2501.ai/0.12/configure/prompting.md): Learn the basics of prompting AI Agents - [Talking to an Agent](https://docs.2501.ai/0.12/prompting/talking-to-agents.md): How to write task prompts that agents actually act on correctly - [Prompting for a Specialty](https://docs.2501.ai/0.12/prompting/specialty.md): Shape how an agent thinks about a domain — vocabulary, priorities, and approach - [Prompting an Operational Rule](https://docs.2501.ai/0.12/prompting/operational-rule.md): Tell the agent what is true about *your* environment - [Prompting a Gateway](https://docs.2501.ai/0.12/prompting/gateway.md): Route and sequence inbound tickets, and govern what 2501 writes back #### Best Practices - [Scoping Agents](https://docs.2501.ai/0.12/best-practices/scoping-agents.md): Why focused agents beat general-purpose ones — and how to scope them - [Working with Knowledge](https://docs.2501.ai/0.12/best-practices/knowledge.md): Feed your existing documentation into 2501 so agents inherit it automatically - [Dynamic Context Injection](https://docs.2501.ai/0.12/best-practices/dynamic-context.md): How 2501 chooses what to put in an agent's context — and what stays out - [MCP & Plugins](https://docs.2501.ai/0.12/best-practices/mcp-plugins.md): When to give agents extra tools — and when not to - [Testing a Behavior](https://docs.2501.ai/0.12/best-practices/testing.md): Verify a new specialty, rule, or MCP before rolling it out to production #### Risk & Safety - [Handling Permissions](https://docs.2501.ai/0.12/risk/permissions.md): Limit agents at the credential and provisioning layer, not just the prompt - [Blacklists in Practice](https://docs.2501.ai/0.12/risk/blacklists.md): The programmatic kill-switch — always on, no matter what - [Read-Only Agents](https://docs.2501.ai/0.12/risk/read-only-agents.md): Use the investigate-only ceiling for critical systems and high-risk new use cases - [Benchmarking for Risk](https://docs.2501.ai/0.12/risk/benchmarking.md): Validate new use cases in a sandbox before exposing them to production #### FAQ - [FAQ](https://docs.2501.ai/0.12/faq.md): Quick answers to common questions about 2501 ## 0.11 ### Docs #### Getting Started - [Welcome](https://docs.2501.ai/0.11/getting-started/welcome.md): AIOps with 2501 — an autonomous agent platform for IT operations - [Quickstart](https://docs.2501.ai/0.11/getting-started/quickstart.md): Install the CLI, deploy 2501, and run your first agent #### Understand 2501 - [Command Center](https://docs.2501.ai/0.11/understand/command-center.md): The web UI for managing your 2501 installation - [Engine & Agents](https://docs.2501.ai/0.11/understand/engine-agents.md): The backend that powers 2501 and the dual-LLM architecture behind every agent - [Agentic Flow](https://docs.2501.ai/0.11/understand/agentic-flow.md): Ticket → gateway → job → tasks → resolution, in one page - [Configuration Approaches](https://docs.2501.ai/0.11/understand/configuration.md): Three ways to configure 2501 — pick the one that fits how you work #### Concepts - [Agents](https://docs.2501.ai/0.11/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.11/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.11/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tickets](https://docs.2501.ai/0.11/core-concepts/tickets.md): Key data extracted from your ITSM, translated for 2501's pipeline - [Tasks](https://docs.2501.ai/0.11/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.11/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Job Schedules](https://docs.2501.ai/0.11/core-concepts/job-schedules.md): Recurring jobs that fan out into a fresh Job each tick - [Gateways](https://docs.2501.ai/0.11/core-concepts/gateways.md): Connect agents to external services ### Reference #### CLI - [2501 CLI](https://docs.2501.ai/0.11/cli/overview.md): One binary, six command groups — manage the platform from your terminal or CI - [login & status](https://docs.2501.ai/0.11/cli/login-status.md): Sign in to your Command Center, confirm who you are - [resources](https://docs.2501.ai/0.11/cli/resources.md): Manage platform resources as version-controlled MDX files - [infra](https://docs.2501.ai/0.11/cli/infra.md): Deploy and operate 2501 on Docker Swarm or Kubernetes - [runner](https://docs.2501.ai/0.11/cli/runner.md): Drive Benchmark scenarios from the CLI - [chat](https://docs.2501.ai/0.11/cli/chat.md): Talk to the AI Assistant from your terminal #### Deploy - [Overview](https://docs.2501.ai/0.11/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.11/deployment/init-command.md): Initialize your 2501 deployment workspace - [Configuration](https://docs.2501.ai/0.11/deployment/configuration.md): How 2501 deployment configuration works - [Docker Swarm](https://docs.2501.ai/0.11/deployment/docker-swarm.md): Production deployment with Docker Swarm - [Kubernetes](https://docs.2501.ai/0.11/deployment/kubernetes.md): Run the 2501 engine on any Kubernetes cluster with the Kubernetes pod executor backend - [LLM Providers](https://docs.2501.ai/0.11/deployment/llm-providers.md): Seed LLM providers at deploy time and manage them in Command Center - [Users & Organizations](https://docs.2501.ai/0.11/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.11/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 - [Feature Toggles](https://docs.2501.ai/0.11/deployment/feature-toggles.md): Toggle experimental and optional behaviors on or off #### Configure - [Knowledge Base](https://docs.2501.ai/0.11/configure/knowledge.md): Upload documents and let the engine extract reusable knowledge for your agents - [Users](https://docs.2501.ai/0.11/configure/users.md): Manage users and permissions from the Command Center - [Licensing & Usage](https://docs.2501.ai/0.11/configure/licensing.md): Apply your 2501 license, understand entitlement caps, and track usage from Command Center ##### Agent Behavior - [Specialties](https://docs.2501.ai/0.11/configure/specialties.md): Configure your agent for specific task domains - [Operational Rules](https://docs.2501.ai/0.11/configure/operational-rules.md): Tag-matched guardrails that apply to the right tasks, not every agent - [Blacklist](https://docs.2501.ai/0.11/configure/blacklist.md): Prevent agents from executing specific commands ##### Access & Secrets - [Credentials](https://docs.2501.ai/0.11/configure/credentials.md): Enable agents to access your services - [Plugins](https://docs.2501.ai/0.11/configure/plugins.md): Connect AI Agents to MCPs and plugins ##### Integration - [Webhooks](https://docs.2501.ai/0.11/configure/webhooks.md): Set up real-time webhook notifications from ServiceNow - [Configuration as Code](https://docs.2501.ai/0.11/configure/configuration-as-code.md): Manage platform resources as version-controlled MDX with the 2501 CLI - [Import Hosts via CSV](https://docs.2501.ai/0.11/configure/import-hosts-csv.md): Bulk import hosts and agents using CSV files - [Network Discovery](https://docs.2501.ai/0.11/configure/discovery.md): Scan your network to build host inventory automatically ###### Resource Reference - [Agent](https://docs.2501.ai/0.11/configure/resources/agents.md): MDX frontmatter reference for agent resources - [Host](https://docs.2501.ai/0.11/configure/resources/hosts.md): MDX frontmatter reference for host resources - [Specialty](https://docs.2501.ai/0.11/configure/resources/specialties.md): MDX frontmatter reference for specialty resources - [Operational Rule](https://docs.2501.ai/0.11/configure/resources/operational-rules.md): MDX frontmatter reference for operational rule resources - [Credential](https://docs.2501.ai/0.11/configure/resources/credentials.md): MDX frontmatter reference for credential resources - [Gateway](https://docs.2501.ai/0.11/configure/resources/gateways.md): MDX frontmatter reference for gateway resources - [Blacklist](https://docs.2501.ai/0.11/configure/resources/blacklist.md): MDX frontmatter reference for blacklist entries - [LLM Provider](https://docs.2501.ai/0.11/configure/resources/providers.md): MDX frontmatter reference for LLM provider resources - [LLM Model](https://docs.2501.ai/0.11/configure/resources/models.md): MDX frontmatter reference for LLM model resources ##### LLM - [Providers](https://docs.2501.ai/0.11/configure/providers.md): Register the LLM endpoints the engine can call - [Models](https://docs.2501.ai/0.11/configure/models.md): Catalog the LLMs your agents can pick as engines ### Benchmark #### Get Started - [Why Benchmark](https://docs.2501.ai/0.11/benchmark/overview.md): Validate agent behavior on realistic scenarios before exposing it to production - [Quickstart](https://docs.2501.ai/0.11/benchmark/quickstart.md): Run your first benchmark scenario end to end #### Authoring - [Scenario Structure](https://docs.2501.ai/0.11/benchmark/scenario.md): Scenario structure, keys, tags, and writing effective descriptions - [Hosts](https://docs.2501.ai/0.11/benchmark/hosts.md): Target host configuration for host mode and VM mode - [Agents](https://docs.2501.ai/0.11/benchmark/agents.md): Agent configuration for host mode and VM mode - [Playbooks](https://docs.2501.ai/0.11/benchmark/playbooks.md): Ansible playbooks for prepare, restore, and validate phases - [Validation](https://docs.2501.ai/0.11/benchmark/validation.md): Validation rules, validators, and scoring model - [Examples](https://docs.2501.ai/0.11/benchmark/examples.md): Complete worked scenarios ready to copy and adapt #### Running - [start](https://docs.2501.ai/0.11/benchmark/start.md): CLI reference for 2501 runner start - [validate](https://docs.2501.ai/0.11/benchmark/validate.md): Lint scenario definitions or re-run validation rules against existing runs - [flush](https://docs.2501.ai/0.11/benchmark/flush.md): Delete scenario run data from the database #### VM Sandbox - [Overview](https://docs.2501.ai/0.11/benchmark/sandbox.md): Isolated, ephemeral VMs for scenario runs and agent testing - [Templates](https://docs.2501.ai/0.11/benchmark/sandbox-templates.md): VM base images available for incus and lima modes - [From a Scenario](https://docs.2501.ai/0.11/benchmark/sandbox-scenario.md): Step into a broken scenario environment to inspect state and iterate on playbooks - [Standalone VMs](https://docs.2501.ai/0.11/benchmark/sandbox-targets.md): Provision standalone VMs for ad-hoc agent testing outside of scenario runs ### Guides #### Prompting - [Prompting](https://docs.2501.ai/0.11/configure/prompting.md): Learn the basics of prompting AI Agents - [Talking to an Agent](https://docs.2501.ai/0.11/prompting/talking-to-agents.md): How to write task prompts that agents actually act on correctly - [Prompting for a Specialty](https://docs.2501.ai/0.11/prompting/specialty.md): Shape how an agent thinks about a domain — vocabulary, priorities, and approach - [Prompting an Operational Rule](https://docs.2501.ai/0.11/prompting/operational-rule.md): Tell the agent what is true about *your* environment - [Prompting a Gateway](https://docs.2501.ai/0.11/prompting/gateway.md): Route and sequence inbound tickets, and govern what 2501 writes back #### Best Practices - [Scoping Agents](https://docs.2501.ai/0.11/best-practices/scoping-agents.md): Why focused agents beat general-purpose ones — and how to scope them - [Working with Knowledge](https://docs.2501.ai/0.11/best-practices/knowledge.md): Feed your existing documentation into 2501 so agents inherit it automatically - [Dynamic Context Injection](https://docs.2501.ai/0.11/best-practices/dynamic-context.md): How 2501 chooses what to put in an agent's context — and what stays out - [MCP & Plugins](https://docs.2501.ai/0.11/best-practices/mcp-plugins.md): When to give agents extra tools — and when not to - [Testing a Behavior](https://docs.2501.ai/0.11/best-practices/testing.md): Verify a new specialty, rule, or MCP before rolling it out to production #### Risk & Safety - [Handling Permissions](https://docs.2501.ai/0.11/risk/permissions.md): Limit agents at the credential and provisioning layer, not just the prompt - [Blacklists in Practice](https://docs.2501.ai/0.11/risk/blacklists.md): The programmatic kill-switch — always on, no matter what - [Read-Only Agents](https://docs.2501.ai/0.11/risk/read-only-agents.md): Use the investigate-only ceiling for critical systems and high-risk new use cases - [Benchmarking for Risk](https://docs.2501.ai/0.11/risk/benchmarking.md): Validate new use cases in a sandbox before exposing them to production #### FAQ - [FAQ](https://docs.2501.ai/0.11/faq.md): Quick answers to common questions about 2501 ## 0.10 ### Docs #### Getting Started - [Welcome](https://docs.2501.ai/0.10/getting-started/welcome.md): AIOps with 2501 — an autonomous agent platform for IT operations - [Quickstart](https://docs.2501.ai/0.10/getting-started/quickstart.md): Install the CLI, deploy 2501, and run your first agent #### Understand 2501 - [Command Center](https://docs.2501.ai/0.10/understand/command-center.md): The web UI for managing your 2501 installation - [Engine & Agents](https://docs.2501.ai/0.10/understand/engine-agents.md): The backend that powers 2501 and the dual-LLM architecture behind every agent - [Agentic Flow](https://docs.2501.ai/0.10/understand/agentic-flow.md): Ticket → gateway → job → tasks → resolution, in one page - [Configuration Approaches](https://docs.2501.ai/0.10/understand/configuration.md): Three ways to configure 2501 — pick the one that fits how you work #### Concepts - [Agents](https://docs.2501.ai/0.10/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.10/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.10/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tickets](https://docs.2501.ai/0.10/core-concepts/tickets.md): Key data extracted from your ITSM, translated for 2501's pipeline - [Tasks](https://docs.2501.ai/0.10/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.10/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Job Schedules](https://docs.2501.ai/0.10/core-concepts/job-schedules.md): Recurring jobs that fan out into a fresh Job each tick - [Gateways](https://docs.2501.ai/0.10/core-concepts/gateways.md): Connect agents to external services ### Reference #### CLI - [2501 CLI](https://docs.2501.ai/0.10/cli/overview.md): One binary, six command groups — manage the platform from your terminal or CI - [login & status](https://docs.2501.ai/0.10/cli/login-status.md): Sign in to your Command Center, confirm who you are - [resources](https://docs.2501.ai/0.10/cli/resources.md): Manage platform resources as version-controlled MDX files - [infra](https://docs.2501.ai/0.10/cli/infra.md): Deploy and operate 2501 on Docker Swarm or Kubernetes - [runner](https://docs.2501.ai/0.10/cli/runner.md): Drive Benchmark scenarios from the CLI - [chat](https://docs.2501.ai/0.10/cli/chat.md): Talk to the AI Assistant from your terminal #### Deploy - [Overview](https://docs.2501.ai/0.10/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.10/deployment/init-command.md): Initialize your 2501 deployment workspace - [Configuration](https://docs.2501.ai/0.10/deployment/configuration.md): How 2501 deployment configuration works - [Docker Swarm](https://docs.2501.ai/0.10/deployment/docker-swarm.md): Production deployment with Docker Swarm - [Kubernetes](https://docs.2501.ai/0.10/deployment/kubernetes.md): Run the 2501 engine on any Kubernetes cluster with the Kubernetes pod executor backend - [LLM Providers](https://docs.2501.ai/0.10/deployment/llm-providers.md): Seed LLM providers at deploy time and manage them in Command Center - [Users & Organizations](https://docs.2501.ai/0.10/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.10/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 - [Feature Toggles](https://docs.2501.ai/0.10/deployment/feature-toggles.md): Toggle experimental and optional behaviors on or off #### Configure - [Knowledge Base](https://docs.2501.ai/0.10/configure/knowledge.md): Upload documents and let the engine extract reusable knowledge for your agents - [Users](https://docs.2501.ai/0.10/configure/users.md): Manage users and permissions from the Command Center - [Licensing & Usage](https://docs.2501.ai/0.10/configure/licensing.md): Apply your 2501 license, understand entitlement caps, and track usage from Command Center ##### Agent Behavior - [Specialties](https://docs.2501.ai/0.10/configure/specialties.md): Configure your agent for specific task domains - [Operational Rules](https://docs.2501.ai/0.10/configure/operational-rules.md): Tag-matched guardrails that apply to the right tasks, not every agent - [Blacklist](https://docs.2501.ai/0.10/configure/blacklist.md): Prevent agents from executing specific commands ##### Access & Secrets - [Credentials](https://docs.2501.ai/0.10/configure/credentials.md): Enable agents to access your services - [Plugins](https://docs.2501.ai/0.10/configure/plugins.md): Connect AI Agents to MCPs and plugins ##### Integration - [Webhooks](https://docs.2501.ai/0.10/configure/webhooks.md): Set up real-time webhook notifications from ServiceNow - [Configuration as Code](https://docs.2501.ai/0.10/configure/configuration-as-code.md): Manage platform resources as version-controlled MDX with the 2501 CLI - [Import Hosts via CSV](https://docs.2501.ai/0.10/configure/import-hosts-csv.md): Bulk import hosts and agents using CSV files - [Network Discovery](https://docs.2501.ai/0.10/configure/discovery.md): Scan your network to build host inventory automatically ###### Resource Reference - [Agent](https://docs.2501.ai/0.10/configure/resources/agents.md): MDX frontmatter reference for agent resources - [Host](https://docs.2501.ai/0.10/configure/resources/hosts.md): MDX frontmatter reference for host resources - [Specialty](https://docs.2501.ai/0.10/configure/resources/specialties.md): MDX frontmatter reference for specialty resources - [Operational Rule](https://docs.2501.ai/0.10/configure/resources/operational-rules.md): MDX frontmatter reference for operational rule resources - [Credential](https://docs.2501.ai/0.10/configure/resources/credentials.md): MDX frontmatter reference for credential resources - [Gateway](https://docs.2501.ai/0.10/configure/resources/gateways.md): MDX frontmatter reference for gateway resources - [Blacklist](https://docs.2501.ai/0.10/configure/resources/blacklist.md): MDX frontmatter reference for blacklist entries ##### LLM - [Providers](https://docs.2501.ai/0.10/configure/providers.md): Register the LLM endpoints the engine can call - [Models](https://docs.2501.ai/0.10/configure/models.md): Catalog the LLMs your agents can pick as engines ### Benchmark #### Get Started - [Why Benchmark](https://docs.2501.ai/0.10/benchmark/overview.md): Validate agent behavior on realistic scenarios before exposing it to production - [Quickstart](https://docs.2501.ai/0.10/benchmark/quickstart.md): Run your first benchmark scenario end to end #### Authoring - [Scenario Structure](https://docs.2501.ai/0.10/benchmark/scenario.md): Scenario structure, keys, tags, and writing effective descriptions - [Hosts](https://docs.2501.ai/0.10/benchmark/hosts.md): Target host configuration for host mode and VM mode - [Agents](https://docs.2501.ai/0.10/benchmark/agents.md): Agent configuration for host mode and VM mode - [Playbooks](https://docs.2501.ai/0.10/benchmark/playbooks.md): Ansible playbooks for prepare, restore, and validate phases - [Validation](https://docs.2501.ai/0.10/benchmark/validation.md): Validation rules, validators, and scoring model - [Examples](https://docs.2501.ai/0.10/benchmark/examples.md): Complete worked scenarios ready to copy and adapt #### Running - [start](https://docs.2501.ai/0.10/benchmark/start.md): CLI reference for 2501 runner start - [validate](https://docs.2501.ai/0.10/benchmark/validate.md): Lint scenario definitions or re-run validation rules against existing runs - [flush](https://docs.2501.ai/0.10/benchmark/flush.md): Delete scenario run data from the database #### VM Sandbox - [Overview](https://docs.2501.ai/0.10/benchmark/sandbox.md): Isolated, ephemeral VMs for scenario runs and agent testing - [Templates](https://docs.2501.ai/0.10/benchmark/sandbox-templates.md): VM base images available for incus and lima modes - [From a Scenario](https://docs.2501.ai/0.10/benchmark/sandbox-scenario.md): Step into a broken scenario environment to inspect state and iterate on playbooks - [Standalone VMs](https://docs.2501.ai/0.10/benchmark/sandbox-targets.md): Provision standalone VMs for ad-hoc agent testing outside of scenario runs ### Guides #### Prompting - [Prompting](https://docs.2501.ai/0.10/configure/prompting.md): Learn the basics of prompting AI Agents - [Talking to an Agent](https://docs.2501.ai/0.10/prompting/talking-to-agents.md): How to write task prompts that agents actually act on correctly - [Prompting for a Specialty](https://docs.2501.ai/0.10/prompting/specialty.md): Shape how an agent thinks about a domain — vocabulary, priorities, and approach - [Prompting an Operational Rule](https://docs.2501.ai/0.10/prompting/operational-rule.md): Tell the agent what is true about *your* environment - [Prompting a Gateway](https://docs.2501.ai/0.10/prompting/gateway.md): Route and sequence inbound tickets, and govern what 2501 writes back #### Best Practices - [Scoping Agents](https://docs.2501.ai/0.10/best-practices/scoping-agents.md): Why focused agents beat general-purpose ones — and how to scope them - [Working with Knowledge](https://docs.2501.ai/0.10/best-practices/knowledge.md): Feed your existing documentation into 2501 so agents inherit it automatically - [Dynamic Context Injection](https://docs.2501.ai/0.10/best-practices/dynamic-context.md): How 2501 chooses what to put in an agent's context — and what stays out - [MCP & Plugins](https://docs.2501.ai/0.10/best-practices/mcp-plugins.md): When to give agents extra tools — and when not to - [Testing a Behavior](https://docs.2501.ai/0.10/best-practices/testing.md): Verify a new specialty, rule, or MCP before rolling it out to production #### Risk & Safety - [Handling Permissions](https://docs.2501.ai/0.10/risk/permissions.md): Limit agents at the credential and provisioning layer, not just the prompt - [Blacklists in Practice](https://docs.2501.ai/0.10/risk/blacklists.md): The programmatic kill-switch — always on, no matter what - [Read-Only Agents](https://docs.2501.ai/0.10/risk/read-only-agents.md): Use the investigate-only ceiling for critical systems and high-risk new use cases - [Benchmarking for Risk](https://docs.2501.ai/0.10/risk/benchmarking.md): Validate new use cases in a sandbox before exposing them to production #### FAQ - [FAQ](https://docs.2501.ai/0.10/faq.md): Quick answers to common questions about 2501 ## 0.9 ### Docs #### Getting Started - [Welcome](https://docs.2501.ai/0.9/getting-started/welcome.md): AIOps with 2501 — an autonomous agent platform for IT operations - [Quickstart](https://docs.2501.ai/0.9/getting-started/quickstart.md): Install the CLI, deploy 2501, and run your first agent #### Understand 2501 - [Command Center](https://docs.2501.ai/0.9/understand/command-center.md): The web UI for managing your 2501 installation - [Engine & Agents](https://docs.2501.ai/0.9/understand/engine-agents.md): The backend that powers 2501 and the dual-LLM architecture behind every agent - [Agentic Flow](https://docs.2501.ai/0.9/understand/agentic-flow.md): Ticket → gateway → job → tasks → resolution, in one page - [Configuration Approaches](https://docs.2501.ai/0.9/understand/configuration.md): Three ways to configure 2501 — pick the one that fits how you work #### Concepts - [Agents](https://docs.2501.ai/0.9/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.9/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.9/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tickets](https://docs.2501.ai/0.9/core-concepts/tickets.md): Key data extracted from your ITSM, translated for 2501's pipeline - [Tasks](https://docs.2501.ai/0.9/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.9/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Job Schedules](https://docs.2501.ai/0.9/core-concepts/job-schedules.md): Recurring jobs that fan out into a fresh Job each tick - [Gateways](https://docs.2501.ai/0.9/core-concepts/gateways.md): Connect agents to external services ### Reference #### CLI - [2501 CLI](https://docs.2501.ai/0.9/cli/overview.md): One binary, six command groups — manage the platform from your terminal or CI - [login & status](https://docs.2501.ai/0.9/cli/login-status.md): Sign in to your Command Center, confirm who you are - [resources](https://docs.2501.ai/0.9/cli/resources.md): Manage platform resources as version-controlled MDX files - [infra](https://docs.2501.ai/0.9/cli/infra.md): Deploy and operate 2501 on Docker Swarm or Kubernetes - [runner](https://docs.2501.ai/0.9/cli/runner.md): Drive Benchmark scenarios from the CLI - [chat](https://docs.2501.ai/0.9/cli/chat.md): Talk to the AI Assistant from your terminal #### Deploy - [Overview](https://docs.2501.ai/0.9/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.9/deployment/init-command.md): Initialize your 2501 deployment workspace - [Configuration](https://docs.2501.ai/0.9/deployment/configuration.md): How 2501 deployment configuration works - [Docker Swarm](https://docs.2501.ai/0.9/deployment/docker-swarm.md): Production deployment with Docker Swarm - [Kubernetes](https://docs.2501.ai/0.9/deployment/kubernetes.md): Run the 2501 engine on any Kubernetes cluster with the Kubernetes pod executor backend - [LLM Providers](https://docs.2501.ai/0.9/deployment/llm-providers.md): Seed LLM providers at deploy time and manage them in Command Center - [Users & Organizations](https://docs.2501.ai/0.9/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.9/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 - [Feature Toggles](https://docs.2501.ai/0.9/deployment/feature-toggles.md): Toggle experimental and optional behaviors on or off #### Configure - [Knowledge Base](https://docs.2501.ai/0.9/configure/knowledge.md): Upload documents and let the engine extract reusable knowledge for your agents - [Users](https://docs.2501.ai/0.9/configure/users.md): Manage users and permissions from the Command Center - [Licensing & Usage](https://docs.2501.ai/0.9/configure/licensing.md): Apply your 2501 license, understand entitlement caps, and track usage from Command Center ##### Agent Behavior - [Specialties](https://docs.2501.ai/0.9/configure/specialties.md): Configure your agent for specific task domains - [Operational Rules](https://docs.2501.ai/0.9/configure/operational-rules.md): Tag-matched guardrails that apply to the right tasks, not every agent - [Blacklist](https://docs.2501.ai/0.9/configure/blacklist.md): Prevent agents from executing specific commands ##### Access & Secrets - [Credentials](https://docs.2501.ai/0.9/configure/credentials.md): Enable agents to access your services - [Plugins](https://docs.2501.ai/0.9/configure/plugins.md): Connect AI Agents to MCPs and plugins ##### Integration - [Webhooks](https://docs.2501.ai/0.9/configure/webhooks.md): Set up real-time webhook notifications from ServiceNow - [Configuration as Code](https://docs.2501.ai/0.9/configure/configuration-as-code.md): Manage platform resources as version-controlled MDX with the 2501 CLI - [Import Hosts via CSV](https://docs.2501.ai/0.9/configure/import-hosts-csv.md): Bulk import hosts and agents using CSV files - [Network Discovery](https://docs.2501.ai/0.9/configure/discovery.md): Scan your network to build host inventory automatically ###### Resource Reference - [Agent](https://docs.2501.ai/0.9/configure/resources/agents.md): MDX frontmatter reference for agent resources - [Host](https://docs.2501.ai/0.9/configure/resources/hosts.md): MDX frontmatter reference for host resources - [Specialty](https://docs.2501.ai/0.9/configure/resources/specialties.md): MDX frontmatter reference for specialty resources - [Operational Rule](https://docs.2501.ai/0.9/configure/resources/operational-rules.md): MDX frontmatter reference for operational rule resources - [Credential](https://docs.2501.ai/0.9/configure/resources/credentials.md): MDX frontmatter reference for credential resources - [Gateway](https://docs.2501.ai/0.9/configure/resources/gateways.md): MDX frontmatter reference for gateway resources - [Blacklist](https://docs.2501.ai/0.9/configure/resources/blacklist.md): MDX frontmatter reference for blacklist entries ##### LLM - [Providers](https://docs.2501.ai/0.9/configure/providers.md): Register the LLM endpoints the engine can call - [Models](https://docs.2501.ai/0.9/configure/models.md): Catalog the LLMs your agents can pick as engines ### Benchmark #### Get Started - [Why Benchmark](https://docs.2501.ai/0.9/benchmark/overview.md): Validate agent behavior on realistic scenarios before exposing it to production - [Quickstart](https://docs.2501.ai/0.9/benchmark/quickstart.md): Run your first benchmark scenario end to end #### Authoring - [Scenario Structure](https://docs.2501.ai/0.9/benchmark/scenario.md): Scenario structure, keys, tags, and writing effective descriptions - [Hosts](https://docs.2501.ai/0.9/benchmark/hosts.md): Target host configuration for host mode and VM mode - [Agents](https://docs.2501.ai/0.9/benchmark/agents.md): Agent configuration for host mode and VM mode - [Playbooks](https://docs.2501.ai/0.9/benchmark/playbooks.md): Ansible playbooks for prepare, restore, and validate phases - [Validation](https://docs.2501.ai/0.9/benchmark/validation.md): Validation rules, validators, and scoring model - [Examples](https://docs.2501.ai/0.9/benchmark/examples.md): Complete worked scenarios ready to copy and adapt #### Running - [start](https://docs.2501.ai/0.9/benchmark/start.md): CLI reference for 2501 runner start - [validate](https://docs.2501.ai/0.9/benchmark/validate.md): Lint scenario definitions or re-run validation rules against existing runs - [flush](https://docs.2501.ai/0.9/benchmark/flush.md): Delete scenario run data from the database #### VM Sandbox - [Overview](https://docs.2501.ai/0.9/benchmark/sandbox.md): Isolated, ephemeral VMs for scenario runs and agent testing - [Templates](https://docs.2501.ai/0.9/benchmark/sandbox-templates.md): VM base images available for incus and lima modes - [From a Scenario](https://docs.2501.ai/0.9/benchmark/sandbox-scenario.md): Step into a broken scenario environment to inspect state and iterate on playbooks - [Standalone VMs](https://docs.2501.ai/0.9/benchmark/sandbox-targets.md): Provision standalone VMs for ad-hoc agent testing outside of scenario runs ### Guides #### Prompting - [Prompting](https://docs.2501.ai/0.9/configure/prompting.md): Learn the basics of prompting AI Agents - [Talking to an Agent](https://docs.2501.ai/0.9/prompting/talking-to-agents.md): How to write task prompts that agents actually act on correctly - [Prompting for a Specialty](https://docs.2501.ai/0.9/prompting/specialty.md): Shape how an agent thinks about a domain — vocabulary, priorities, and approach - [Prompting an Operational Rule](https://docs.2501.ai/0.9/prompting/operational-rule.md): Tell the agent what is true about *your* environment - [Prompting a Gateway](https://docs.2501.ai/0.9/prompting/gateway.md): Route and sequence inbound tickets, and govern what 2501 writes back #### Best Practices - [Scoping Agents](https://docs.2501.ai/0.9/best-practices/scoping-agents.md): Why focused agents beat general-purpose ones — and how to scope them - [Working with Knowledge](https://docs.2501.ai/0.9/best-practices/knowledge.md): Feed your existing documentation into 2501 so agents inherit it automatically - [Dynamic Context Injection](https://docs.2501.ai/0.9/best-practices/dynamic-context.md): How 2501 chooses what to put in an agent's context — and what stays out - [MCP & Plugins](https://docs.2501.ai/0.9/best-practices/mcp-plugins.md): When to give agents extra tools — and when not to - [Testing a Behavior](https://docs.2501.ai/0.9/best-practices/testing.md): Verify a new specialty, rule, or MCP before rolling it out to production #### Risk & Safety - [Handling Permissions](https://docs.2501.ai/0.9/risk/permissions.md): Limit agents at the credential and provisioning layer, not just the prompt - [Blacklists in Practice](https://docs.2501.ai/0.9/risk/blacklists.md): The programmatic kill-switch — always on, no matter what - [Read-Only Agents](https://docs.2501.ai/0.9/risk/read-only-agents.md): Use the investigate-only ceiling for critical systems and high-risk new use cases - [Benchmarking for Risk](https://docs.2501.ai/0.9/risk/benchmarking.md): Validate new use cases in a sandbox before exposing them to production #### FAQ - [FAQ](https://docs.2501.ai/0.9/faq.md): Quick answers to common questions about 2501 ## 0.8 ### Docs #### Getting Started - [Welcome](https://docs.2501.ai/0.8/getting-started/welcome.md): AIOps with 2501 — an autonomous agent platform for IT operations - [Quickstart](https://docs.2501.ai/0.8/getting-started/quickstart.md): Install the CLI, deploy 2501, and run your first agent #### Understand 2501 - [Command Center](https://docs.2501.ai/0.8/understand/command-center.md): The web UI for managing your 2501 installation - [Engine & Agents](https://docs.2501.ai/0.8/understand/engine-agents.md): The backend that powers 2501 and the dual-LLM architecture behind every agent - [Agentic Flow](https://docs.2501.ai/0.8/understand/agentic-flow.md): Ticket → gateway → job → tasks → resolution, in one page - [Configuration Approaches](https://docs.2501.ai/0.8/understand/configuration.md): Three ways to configure 2501 — pick the one that fits how you work #### Concepts - [Agents](https://docs.2501.ai/0.8/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.8/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.8/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tickets](https://docs.2501.ai/0.8/core-concepts/tickets.md): Key data extracted from your ITSM, translated for 2501's pipeline - [Tasks](https://docs.2501.ai/0.8/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.8/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Job Schedules](https://docs.2501.ai/0.8/core-concepts/job-schedules.md): Recurring jobs that fan out into a fresh Job each tick - [Gateways](https://docs.2501.ai/0.8/core-concepts/gateways.md): Connect agents to external services ### Reference #### CLI - [2501 CLI](https://docs.2501.ai/0.8/cli/overview.md): One binary, six command groups — manage the platform from your terminal or CI - [login & status](https://docs.2501.ai/0.8/cli/login-status.md): Sign in to your Command Center, confirm who you are - [resources](https://docs.2501.ai/0.8/cli/resources.md): Manage platform resources as version-controlled MDX files - [infra](https://docs.2501.ai/0.8/cli/infra.md): Deploy and operate 2501 on Docker Swarm or Kubernetes - [runner](https://docs.2501.ai/0.8/cli/runner.md): Drive Benchmark scenarios from the CLI - [chat](https://docs.2501.ai/0.8/cli/chat.md): Talk to the AI Assistant from your terminal #### Deploy - [Overview](https://docs.2501.ai/0.8/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.8/deployment/init-command.md): Initialize your 2501 deployment workspace - [Configuration](https://docs.2501.ai/0.8/deployment/configuration.md): How 2501 deployment configuration works - [Docker Swarm](https://docs.2501.ai/0.8/deployment/docker-swarm.md): Production deployment with Docker Swarm - [Kubernetes](https://docs.2501.ai/0.8/deployment/kubernetes.md): Run the 2501 engine on any Kubernetes cluster with the Kubernetes pod executor backend - [LLM Providers](https://docs.2501.ai/0.8/deployment/llm-providers.md): Seed LLM providers at deploy time and manage them in Command Center - [Users & Organizations](https://docs.2501.ai/0.8/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.8/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 - [Feature Toggles](https://docs.2501.ai/0.8/deployment/feature-toggles.md): Toggle experimental and optional behaviors on or off #### Configure - [Knowledge Base](https://docs.2501.ai/0.8/configure/knowledge.md): Upload documents and let the engine extract reusable knowledge for your agents - [Users](https://docs.2501.ai/0.8/configure/users.md): Manage users and permissions from the Command Center - [Licensing & Usage](https://docs.2501.ai/0.8/configure/licensing.md): Apply your 2501 license, understand entitlement caps, and track usage from Command Center ##### Agent Behavior - [Specialties](https://docs.2501.ai/0.8/configure/specialties.md): Configure your agent for specific task domains - [Operational Rules](https://docs.2501.ai/0.8/configure/operational-rules.md): Tag-matched guardrails that apply to the right tasks, not every agent - [Blacklist](https://docs.2501.ai/0.8/configure/blacklist.md): Prevent agents from executing specific commands ##### Access & Secrets - [Credentials](https://docs.2501.ai/0.8/configure/credentials.md): Enable agents to access your services - [Plugins](https://docs.2501.ai/0.8/configure/plugins.md): Connect AI Agents to MCPs and plugins ##### Integration - [Webhooks](https://docs.2501.ai/0.8/configure/webhooks.md): Set up real-time webhook notifications from ServiceNow - [Configuration as Code](https://docs.2501.ai/0.8/configure/configuration-as-code.md): Manage platform resources as version-controlled MDX with the 2501 CLI - [Import Hosts via CSV](https://docs.2501.ai/0.8/configure/import-hosts-csv.md): Bulk import hosts and agents using CSV files ###### Resource Reference - [Agent](https://docs.2501.ai/0.8/configure/resources/agents.md): MDX frontmatter reference for agent resources - [Host](https://docs.2501.ai/0.8/configure/resources/hosts.md): MDX frontmatter reference for host resources - [Specialty](https://docs.2501.ai/0.8/configure/resources/specialties.md): MDX frontmatter reference for specialty resources - [Operational Rule](https://docs.2501.ai/0.8/configure/resources/operational-rules.md): MDX frontmatter reference for operational rule resources - [Credential](https://docs.2501.ai/0.8/configure/resources/credentials.md): MDX frontmatter reference for credential resources - [Gateway](https://docs.2501.ai/0.8/configure/resources/gateways.md): MDX frontmatter reference for gateway resources - [Blacklist](https://docs.2501.ai/0.8/configure/resources/blacklist.md): MDX frontmatter reference for blacklist entries ##### LLM - [Providers](https://docs.2501.ai/0.8/configure/providers.md): Register the LLM endpoints the engine can call - [Models](https://docs.2501.ai/0.8/configure/models.md): Catalog the LLMs your agents can pick as engines ### Benchmark #### Get Started - [Why Benchmark](https://docs.2501.ai/0.8/benchmark/overview.md): Validate agent behavior on realistic scenarios before exposing it to production - [Quickstart](https://docs.2501.ai/0.8/benchmark/quickstart.md): Run your first benchmark scenario end to end #### Authoring - [Scenario Structure](https://docs.2501.ai/0.8/benchmark/scenario.md): Scenario structure, keys, tags, and writing effective descriptions - [Hosts](https://docs.2501.ai/0.8/benchmark/hosts.md): Target host configuration for host mode and VM mode - [Agents](https://docs.2501.ai/0.8/benchmark/agents.md): Agent configuration for host mode and VM mode - [Playbooks](https://docs.2501.ai/0.8/benchmark/playbooks.md): Ansible playbooks for prepare, restore, and validate phases - [Validation](https://docs.2501.ai/0.8/benchmark/validation.md): Validation rules, validators, and scoring model - [Examples](https://docs.2501.ai/0.8/benchmark/examples.md): Complete worked scenarios ready to copy and adapt #### Running - [start](https://docs.2501.ai/0.8/benchmark/start.md): CLI reference for 2501 runner start - [validate](https://docs.2501.ai/0.8/benchmark/validate.md): Lint scenario definitions or re-run validation rules against existing runs - [flush](https://docs.2501.ai/0.8/benchmark/flush.md): Delete scenario run data from the database #### VM Sandbox - [Overview](https://docs.2501.ai/0.8/benchmark/sandbox.md): Isolated, ephemeral VMs for scenario runs and agent testing - [Templates](https://docs.2501.ai/0.8/benchmark/sandbox-templates.md): VM base images available for incus and lima modes - [From a Scenario](https://docs.2501.ai/0.8/benchmark/sandbox-scenario.md): Step into a broken scenario environment to inspect state and iterate on playbooks - [Standalone VMs](https://docs.2501.ai/0.8/benchmark/sandbox-targets.md): Provision standalone VMs for ad-hoc agent testing outside of scenario runs ### Guides #### Prompting - [Prompting](https://docs.2501.ai/0.8/configure/prompting.md): Learn the basics of prompting AI Agents - [Talking to an Agent](https://docs.2501.ai/0.8/prompting/talking-to-agents.md): How to write task prompts that agents actually act on correctly - [Prompting for a Specialty](https://docs.2501.ai/0.8/prompting/specialty.md): Shape how an agent thinks about a domain — vocabulary, priorities, and approach - [Prompting an Operational Rule](https://docs.2501.ai/0.8/prompting/operational-rule.md): Tell the agent what is true about *your* environment - [Prompting a Gateway](https://docs.2501.ai/0.8/prompting/gateway.md): Route and sequence inbound tickets, and govern what 2501 writes back #### Best Practices - [Scoping Agents](https://docs.2501.ai/0.8/best-practices/scoping-agents.md): Why focused agents beat general-purpose ones — and how to scope them - [Working with Knowledge](https://docs.2501.ai/0.8/best-practices/knowledge.md): Feed your existing documentation into 2501 so agents inherit it automatically - [Dynamic Context Injection](https://docs.2501.ai/0.8/best-practices/dynamic-context.md): How 2501 chooses what to put in an agent's context — and what stays out - [MCP & Plugins](https://docs.2501.ai/0.8/best-practices/mcp-plugins.md): When to give agents extra tools — and when not to - [Testing a Behavior](https://docs.2501.ai/0.8/best-practices/testing.md): Verify a new specialty, rule, or MCP before rolling it out to production #### Risk & Safety - [Handling Permissions](https://docs.2501.ai/0.8/risk/permissions.md): Limit agents at the credential and provisioning layer, not just the prompt - [Blacklists in Practice](https://docs.2501.ai/0.8/risk/blacklists.md): The programmatic kill-switch — always on, no matter what - [Read-Only Agents](https://docs.2501.ai/0.8/risk/read-only-agents.md): Use the investigate-only ceiling for critical systems and high-risk new use cases - [Benchmarking for Risk](https://docs.2501.ai/0.8/risk/benchmarking.md): Validate new use cases in a sandbox before exposing them to production #### FAQ - [FAQ](https://docs.2501.ai/0.8/faq.md): Quick answers to common questions about 2501 ## 0.7 ### Docs #### Getting Started - [Welcome](https://docs.2501.ai/0.7/getting-started/welcome.md): AIOps with 2501 — an autonomous agent platform for IT operations - [Quickstart](https://docs.2501.ai/0.7/getting-started/quickstart.md): Install the CLI, deploy 2501, and run your first agent #### Understand 2501 - [Command Center](https://docs.2501.ai/0.7/understand/command-center.md): The web UI for managing your 2501 installation - [Engine & Agents](https://docs.2501.ai/0.7/understand/engine-agents.md): The backend that powers 2501 and the dual-LLM architecture behind every agent - [Agentic Flow](https://docs.2501.ai/0.7/understand/agentic-flow.md): Ticket → gateway → job → tasks → resolution, in one page - [Configuration Approaches](https://docs.2501.ai/0.7/understand/configuration.md): Three ways to configure 2501 — pick the one that fits how you work #### Concepts - [Agents](https://docs.2501.ai/0.7/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.7/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.7/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tickets](https://docs.2501.ai/0.7/core-concepts/tickets.md): Key data extracted from your ITSM, translated for 2501's pipeline - [Tasks](https://docs.2501.ai/0.7/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.7/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Job Schedules](https://docs.2501.ai/0.7/core-concepts/job-schedules.md): Recurring jobs that fan out into a fresh Job each tick - [Gateways](https://docs.2501.ai/0.7/core-concepts/gateways.md): Connect agents to external services ### Reference #### CLI - [2501 CLI](https://docs.2501.ai/0.7/cli/overview.md): One binary, six command groups — manage the platform from your terminal or CI - [login & status](https://docs.2501.ai/0.7/cli/login-status.md): Sign in to your Command Center, confirm who you are - [resources](https://docs.2501.ai/0.7/cli/resources.md): Manage platform resources as version-controlled MDX files - [infra](https://docs.2501.ai/0.7/cli/infra.md): Deploy and operate 2501 on Docker Swarm or Kubernetes - [runner](https://docs.2501.ai/0.7/cli/runner.md): Drive Benchmark scenarios from the CLI - [chat](https://docs.2501.ai/0.7/cli/chat.md): Talk to the AI Assistant from your terminal #### Deploy - [Overview](https://docs.2501.ai/0.7/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.7/deployment/init-command.md): Initialize your 2501 deployment workspace - [Configuration](https://docs.2501.ai/0.7/deployment/configuration.md): How 2501 deployment configuration works - [Docker Swarm](https://docs.2501.ai/0.7/deployment/docker-swarm.md): Production deployment with Docker Swarm - [Kubernetes](https://docs.2501.ai/0.7/deployment/kubernetes.md): Run the 2501 engine on any Kubernetes cluster with the Kubernetes pod executor backend - [LLM Providers](https://docs.2501.ai/0.7/deployment/llm-providers.md): Seed LLM providers at deploy time and manage them in Command Center - [Users & Organizations](https://docs.2501.ai/0.7/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.7/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 - [Feature Toggles](https://docs.2501.ai/0.7/deployment/feature-toggles.md): Toggle experimental and optional behaviors on or off #### Configure - [Knowledge Base](https://docs.2501.ai/0.7/configure/knowledge.md): Upload documents and let the engine extract reusable knowledge for your agents - [Users](https://docs.2501.ai/0.7/configure/users.md): Manage users and permissions from the Command Center - [Licensing & Usage](https://docs.2501.ai/0.7/configure/licensing.md): Apply your 2501 license, understand entitlement caps, and track usage from Command Center ##### Agent Behavior - [Specialties](https://docs.2501.ai/0.7/configure/specialties.md): Configure your agent for specific task domains - [Operational Rules](https://docs.2501.ai/0.7/configure/operational-rules.md): Tag-matched guardrails that apply to the right tasks, not every agent - [Blacklist](https://docs.2501.ai/0.7/configure/blacklist.md): Prevent agents from executing specific commands ##### Access & Secrets - [Credentials](https://docs.2501.ai/0.7/configure/credentials.md): Enable agents to access your services - [Plugins](https://docs.2501.ai/0.7/configure/plugins.md): Connect AI Agents to MCPs and plugins ##### Integration - [Webhooks](https://docs.2501.ai/0.7/configure/webhooks.md): Set up real-time webhook notifications from ServiceNow - [Configuration as Code](https://docs.2501.ai/0.7/configure/configuration-as-code.md): Manage platform resources as version-controlled MDX with the 2501 CLI - [Import Hosts via CSV](https://docs.2501.ai/0.7/configure/import-hosts-csv.md): Bulk import hosts and agents using CSV files ###### Resource Reference - [Agent](https://docs.2501.ai/0.7/configure/resources/agents.md): MDX frontmatter reference for agent resources - [Host](https://docs.2501.ai/0.7/configure/resources/hosts.md): MDX frontmatter reference for host resources - [Specialty](https://docs.2501.ai/0.7/configure/resources/specialties.md): MDX frontmatter reference for specialty resources - [Operational Rule](https://docs.2501.ai/0.7/configure/resources/operational-rules.md): MDX frontmatter reference for operational rule resources - [Credential](https://docs.2501.ai/0.7/configure/resources/credentials.md): MDX frontmatter reference for credential resources - [Gateway](https://docs.2501.ai/0.7/configure/resources/gateways.md): MDX frontmatter reference for gateway resources - [Blacklist](https://docs.2501.ai/0.7/configure/resources/blacklist.md): MDX frontmatter reference for blacklist entries ##### LLM - [Providers](https://docs.2501.ai/0.7/configure/providers.md): Register the LLM endpoints the engine can call - [Models](https://docs.2501.ai/0.7/configure/models.md): Catalog the LLMs your agents can pick as engines ### Benchmark #### Get Started - [Why Benchmark](https://docs.2501.ai/0.7/benchmark/overview.md): Validate agent behavior on realistic scenarios before exposing it to production - [Quickstart](https://docs.2501.ai/0.7/benchmark/quickstart.md): Run your first benchmark scenario end to end #### Authoring - [Scenario Structure](https://docs.2501.ai/0.7/benchmark/scenario.md): Scenario structure, keys, tags, and writing effective descriptions - [Hosts](https://docs.2501.ai/0.7/benchmark/hosts.md): Target host configuration for host mode and VM mode - [Agents](https://docs.2501.ai/0.7/benchmark/agents.md): Agent configuration for host mode and VM mode - [Playbooks](https://docs.2501.ai/0.7/benchmark/playbooks.md): Ansible playbooks for prepare, restore, and validate phases - [Validation](https://docs.2501.ai/0.7/benchmark/validation.md): Validation rules, validators, and scoring model - [Examples](https://docs.2501.ai/0.7/benchmark/examples.md): Complete worked scenarios ready to copy and adapt #### Running - [start](https://docs.2501.ai/0.7/benchmark/start.md): CLI reference for 2501 runner start - [validate](https://docs.2501.ai/0.7/benchmark/validate.md): Lint scenario definitions or re-run validation rules against existing runs - [flush](https://docs.2501.ai/0.7/benchmark/flush.md): Delete scenario run data from the database #### VM Sandbox - [Overview](https://docs.2501.ai/0.7/benchmark/sandbox.md): Isolated, ephemeral VMs for scenario runs and agent testing - [Templates](https://docs.2501.ai/0.7/benchmark/sandbox-templates.md): VM base images available for incus and lima modes - [From a Scenario](https://docs.2501.ai/0.7/benchmark/sandbox-scenario.md): Step into a broken scenario environment to inspect state and iterate on playbooks - [Standalone VMs](https://docs.2501.ai/0.7/benchmark/sandbox-targets.md): Provision standalone VMs for ad-hoc agent testing outside of scenario runs ### Guides #### Prompting - [Prompting](https://docs.2501.ai/0.7/configure/prompting.md): Learn the basics of prompting AI Agents - [Talking to an Agent](https://docs.2501.ai/0.7/prompting/talking-to-agents.md): How to write task prompts that agents actually act on correctly - [Prompting for a Specialty](https://docs.2501.ai/0.7/prompting/specialty.md): Shape how an agent thinks about a domain — vocabulary, priorities, and approach - [Prompting an Operational Rule](https://docs.2501.ai/0.7/prompting/operational-rule.md): Tell the agent what is true about *your* environment - [Prompting a Gateway](https://docs.2501.ai/0.7/prompting/gateway.md): Route and sequence — how tickets become the right tasks on the right hosts #### Best Practices - [Scoping Agents](https://docs.2501.ai/0.7/best-practices/scoping-agents.md): Why focused agents beat general-purpose ones — and how to scope them - [Working with Knowledge](https://docs.2501.ai/0.7/best-practices/knowledge.md): Feed your existing documentation into 2501 so agents inherit it automatically - [Dynamic Context Injection](https://docs.2501.ai/0.7/best-practices/dynamic-context.md): How 2501 chooses what to put in an agent's context — and what stays out - [MCP & Plugins](https://docs.2501.ai/0.7/best-practices/mcp-plugins.md): When to give agents extra tools — and when not to - [Testing a Behavior](https://docs.2501.ai/0.7/best-practices/testing.md): Verify a new specialty, rule, or MCP before rolling it out to production #### Risk & Safety - [Handling Permissions](https://docs.2501.ai/0.7/risk/permissions.md): Limit agents at the credential and provisioning layer, not just the prompt - [Blacklists in Practice](https://docs.2501.ai/0.7/risk/blacklists.md): The programmatic kill-switch — always on, no matter what - [Read-Only Agents](https://docs.2501.ai/0.7/risk/read-only-agents.md): Use the investigate-only ceiling for critical systems and high-risk new use cases - [Benchmarking for Risk](https://docs.2501.ai/0.7/risk/benchmarking.md): Validate new use cases in a sandbox before exposing them to production #### FAQ - [FAQ](https://docs.2501.ai/0.7/faq.md): Quick answers to common questions about 2501 ## 0.4 ### Documentation #### Getting Started - [Command Center](https://docs.2501.ai/0.4/command-center.md): The command center for your 2501 application #### Deployment - [Overview](https://docs.2501.ai/0.4/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.4/deployment/init-command.md): Initialize your 2501 deployment configuration - [Configuration](https://docs.2501.ai/0.4/deployment/configuration.md): Overview of 2501 configuration files - [Docker Swarm](https://docs.2501.ai/0.4/deployment/docker-swarm.md): Production deployment with Docker Swarm - [LLM Providers](https://docs.2501.ai/0.4/deployment/llm-providers.md): Configure LLM providers for 2501 - [Users & Organizations](https://docs.2501.ai/0.4/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.4/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 #### Core Concepts - [Agents](https://docs.2501.ai/0.4/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.4/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.4/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tasks](https://docs.2501.ai/0.4/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.4/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Gateways](https://docs.2501.ai/0.4/core-concepts/gateways.md): Connect agents to external services #### Configure - [Specialties](https://docs.2501.ai/0.4/configure/specialties.md): Configure your agent for specific task domains - [Engines](https://docs.2501.ai/0.4/configure/engines.md): Understand how your agents are powered - [Operational Rules](https://docs.2501.ai/0.4/configure/operational-rules.md): Define organization-wide directives for agent behavior - [Blacklist](https://docs.2501.ai/0.4/configure/blacklist.md): Prevent agents from executing specific commands - [Credentials](https://docs.2501.ai/0.4/configure/credentials.md): Enable agents to access your services - [Plugins](https://docs.2501.ai/0.4/configure/plugins.md): Connect AI Agents to MCPs and plugins - [Users](https://docs.2501.ai/0.4/configure/users.md): Manage users and permissions from the Command Center - [Prompting](https://docs.2501.ai/0.4/configure/prompting.md): Learn the basics of prompting AI Agents - [Import Hosts via CSV](https://docs.2501.ai/0.4/configure/import-hosts-csv.md): Bulk import hosts and agents using CSV files - [Webhooks](https://docs.2501.ai/0.4/configure/webhooks.md): Set up real-time webhook notifications from ServiceNow #### Scenario Runner - [Overview](https://docs.2501.ai/0.4/scenario-runner/overview.md): Benchmark 2501 agents against realistic infrastructure scenarios in a sandbox environment ##### Scenarios - [scenario.json](https://docs.2501.ai/0.4/scenario-runner/scenario.md): Scenario structure, keys, tags, and writing effective descriptions - [Hosts](https://docs.2501.ai/0.4/scenario-runner/hosts.md): Target host configuration for host mode and VM mode - [Agents](https://docs.2501.ai/0.4/scenario-runner/agents.md): Agent configuration for host mode and VM mode - [Validation](https://docs.2501.ai/0.4/scenario-runner/validation.md): Validation rules, validators, and scoring model - [Playbooks](https://docs.2501.ai/0.4/scenario-runner/playbooks.md): Ansible playbooks for prepare, restore, and validate phases - [Examples](https://docs.2501.ai/0.4/scenario-runner/examples.md): Complete worked scenarios ready to copy and adapt ##### Commands - [run](https://docs.2501.ai/0.4/scenario-runner/run.md): CLI reference for 2501-runner run - [validate](https://docs.2501.ai/0.4/scenario-runner/validate.md): Lint scenario definitions or re-run validation rules against existing runs - [flush](https://docs.2501.ai/0.4/scenario-runner/flush.md): Delete scenario run data from the database ##### VM Sandbox - [VM Sandbox](https://docs.2501.ai/0.4/scenario-runner/sandbox.md): Isolated, ephemeral VMs for scenario runs and agent testing - [Templates](https://docs.2501.ai/0.4/scenario-runner/sandbox-templates.md): VM base images available for incus and lima modes - [VM from Scenario](https://docs.2501.ai/0.4/scenario-runner/sandbox-scenario.md): Step into a broken scenario environment to inspect state and iterate on playbooks - [VM for Testing](https://docs.2501.ai/0.4/scenario-runner/sandbox-targets.md): Provision standalone VMs for ad-hoc agent testing outside of scenario runs #### CLI - [Installation](https://docs.2501.ai/0.4/cli/installation.md): How to install 2501 CLI on your machine - [Authentication](https://docs.2501.ai/0.4/cli/authentication.md): Authenticate to the CLI by creating secure API Keys - [Commands](https://docs.2501.ai/0.4/cli/commands.md): Discover all the available commands in 2501 CLI ## 0.3.0 ### Documentation #### Getting Started - [Command Center](https://docs.2501.ai/0.3.0/command-center.md): The command center for your 2501 application #### Deployment - [Overview](https://docs.2501.ai/0.3.0/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.3.0/deployment/init-command.md): Initialize your 2501 deployment configuration - [Configuration](https://docs.2501.ai/0.3.0/deployment/configuration.md): Overview of 2501 configuration files - [Docker Swarm](https://docs.2501.ai/0.3.0/deployment/docker-swarm.md): Production deployment with Docker Swarm - [LLM Providers](https://docs.2501.ai/0.3.0/deployment/llm-providers.md): Configure LLM providers for 2501 - [Users & Organizations](https://docs.2501.ai/0.3.0/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.3.0/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 #### Core Concepts - [Agents](https://docs.2501.ai/0.3.0/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.3.0/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.3.0/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tasks](https://docs.2501.ai/0.3.0/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.3.0/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Gateways](https://docs.2501.ai/0.3.0/core-concepts/gateways.md): Connect agents to external services #### Configure - [Specialties](https://docs.2501.ai/0.3.0/configure/specialties.md): Configure your agent for specific task domains - [Engines](https://docs.2501.ai/0.3.0/configure/engines.md): Understand how your agents are powered - [Operational Rules](https://docs.2501.ai/0.3.0/configure/operational-rules.md): Define organization-wide directives for agent behavior - [Blacklist](https://docs.2501.ai/0.3.0/configure/blacklist.md): Prevent agents from executing specific commands - [Credentials](https://docs.2501.ai/0.3.0/configure/credentials.md): Enable agents to access your services - [Plugins](https://docs.2501.ai/0.3.0/configure/plugins.md): Connect AI Agents to MCPs and plugins - [Users](https://docs.2501.ai/0.3.0/configure/users.md): Manage users and permissions from the Command Center - [Prompting](https://docs.2501.ai/0.3.0/configure/prompting.md): Learn the basics of prompting AI Agents - [Import Hosts via CSV](https://docs.2501.ai/0.3.0/configure/import-hosts-csv.md): Bulk import hosts and agents using CSV files #### Scenario Runner - [Overview](https://docs.2501.ai/0.3.0/scenario-runner/overview.md): Benchmark 2501 agents against realistic infrastructure scenarios in a sandbox environment - [Scenario Structure](https://docs.2501.ai/0.3.0/scenario-runner/scenario-structure.md): Directory layout and the scenario.json format - [Hosts & Agents](https://docs.2501.ai/0.3.0/scenario-runner/hosts-and-agents.md): Referencing your sandbox hosts and agents in scenario.json - [Validation](https://docs.2501.ai/0.3.0/scenario-runner/validation.md): Validation rules, validators, scoring model, and fail-fast guards - [Playbooks](https://docs.2501.ai/0.3.0/scenario-runner/playbooks.md): prepare.yml, restore.yml, validate.yml, and Ansible inventory - [Examples](https://docs.2501.ai/0.3.0/scenario-runner/examples.md): Complete scenario examples: disk full, nginx broken config, Kubernetes crashloop - [Run Scenarios](https://docs.2501.ai/0.3.0/scenario-runner/run-command.md): CLI reference, execution flow, entry points, environment setup, and CI integration #### CLI - [Installation](https://docs.2501.ai/0.3.0/cli/installation.md): How to install 2501 CLI on your machine - [Authentication](https://docs.2501.ai/0.3.0/cli/authentication.md): Authenticate to the CLI by creating secure API Keys - [Commands](https://docs.2501.ai/0.3.0/cli/commands.md): Discover all the available commands in 2501 CLI ## 0.2.6 ### Documentation #### Getting Started - [Command Center](https://docs.2501.ai/0.2.6/command-center.md): The command center for your 2501 application #### Deployment - [Overview](https://docs.2501.ai/0.2.6/deployment/overview.md): Deploy 2501 on your infrastructure - [Init Command](https://docs.2501.ai/0.2.6/deployment/init-command.md): Initialize your 2501 deployment configuration - [Configuration](https://docs.2501.ai/0.2.6/deployment/configuration.md): Overview of 2501 configuration files - [Docker Swarm](https://docs.2501.ai/0.2.6/deployment/docker-swarm.md): Production deployment with Docker Swarm - [LLM Providers](https://docs.2501.ai/0.2.6/deployment/llm-providers.md): Configure LLM providers for 2501 - [Users & Organizations](https://docs.2501.ai/0.2.6/deployment/users-organizations.md): Manage users, organizations, and tenants in 2501 - [Troubleshooting](https://docs.2501.ai/0.2.6/deployment/troubleshooting.md): Troubleshooting guide and feature toggles reference for 2501 #### Core Concepts - [Agents](https://docs.2501.ai/0.2.6/core-concepts/agents.md): The LLMs autonomously browsing your systems - [Hosts](https://docs.2501.ai/0.2.6/core-concepts/hosts.md): Your machines where agents act autonomously - [Organizations](https://docs.2501.ai/0.2.6/core-concepts/organizations.md): Organize your tenant by separating between organizations - [Tasks](https://docs.2501.ai/0.2.6/core-concepts/tasks.md): Assign complex tasks to your agents in various ways - [Jobs](https://docs.2501.ai/0.2.6/core-concepts/jobs.md): Coordinate multiple tasks across agents - [Gateways](https://docs.2501.ai/0.2.6/core-concepts/gateways.md): Connect agents to external services #### Configure - [Specialties](https://docs.2501.ai/0.2.6/configure/specialties.md): Configure your agent for specific task domains - [Engines](https://docs.2501.ai/0.2.6/configure/engines.md): Understand how your agents are powered - [Operational Rules](https://docs.2501.ai/0.2.6/configure/operational-rules.md): Define organization-wide directives for agent behavior - [Blacklist](https://docs.2501.ai/0.2.6/configure/blacklist.md): Prevent agents from executing specific commands - [Credentials](https://docs.2501.ai/0.2.6/configure/credentials.md): Enable agents to access your services - [Plugins](https://docs.2501.ai/0.2.6/configure/plugins.md): Connect AI Agents to MCPs and plugins - [Users](https://docs.2501.ai/0.2.6/configure/users.md): Manage users and permissions from the Command Center - [Prompting](https://docs.2501.ai/0.2.6/configure/prompting.md): Learn the basics of prompting AI Agents - [Import Hosts via CSV](https://docs.2501.ai/0.2.6/configure/import-hosts-csv.md): Bulk import hosts and agents using CSV files #### CLI - [Installation](https://docs.2501.ai/0.2.6/cli/installation.md): How to install 2501 CLI on your machine - [Authentication](https://docs.2501.ai/0.2.6/cli/authentication.md): Authenticate to the CLI by creating secure API Keys - [Commands](https://docs.2501.ai/0.2.6/cli/commands.md): Discover all the available commands in 2501 CLI ## OpenAPI Specs - [openapi](/0.20/openapi.json) - [openapi](/0.19/openapi.json) - [openapi](/0.18/openapi.json) - [openapi](/0.17/openapi.json) - [openapi](/0.16/openapi.json) - [openapi](/0.15/openapi.json) - [openapi](/0.14/openapi.json) - [openapi](/0.13/openapi.json) - [openapi](/0.12/openapi.json) - [openapi](/next-release/openapi.json) > The links below point to documentation indexes. Follow each `/_llms/` index recursively until you reach documentation pages. ## Indexes - [0.20 (176 pages)](https://docs.2501.ai/_llms/0-20.txt): Docs, Reference, Benchmark, Guides. - [0.20 / Reference (135 pages)](https://docs.2501.ai/_llms/0-20/reference.txt): API, Deploy, Configure. - [0.19 (164 pages)](https://docs.2501.ai/_llms/0-19.txt): Docs, Reference, Benchmark, Guides. - [0.19 / Reference (123 pages)](https://docs.2501.ai/_llms/0-19/reference.txt): API, Deploy, Configure. - [0.18 (161 pages)](https://docs.2501.ai/_llms/0-18.txt): Docs, Reference, Benchmark, Guides. - [0.18 / Reference (118 pages)](https://docs.2501.ai/_llms/0-18/reference.txt): API, CLI, Deploy, Configure. - [0.17 (167 pages)](https://docs.2501.ai/_llms/0-17.txt): Docs, Reference, Benchmark, Guides. - [0.17 / Reference (127 pages)](https://docs.2501.ai/_llms/0-17/reference.txt): API, CLI, Deploy, Configure. - [0.16 (143 pages)](https://docs.2501.ai/_llms/0-16.txt): Docs, Reference, Benchmark, Guides. - [0.15 (140 pages)](https://docs.2501.ai/_llms/0-15.txt): Docs, Reference, Benchmark, Guides. - [0.14 (126 pages)](https://docs.2501.ai/_llms/0-14.txt): Docs, Reference, Benchmark, Guides. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.