Decision value, not equal-depth directory rows.
Tier 1 AI-assisted launch evidence records are prioritized. Lower-depth records are labeled clearly and default to evidence hold unless claim-level evidence supports a stronger verdict.
AiderTerminal agentsAI-assisted launch evidence recordConditional pilotshell command executionOpen-source terminal editing can be safe to pilot in a sandbox but shell/local repo context remains organization-controlled risk.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Use Aider in a local sandbox or disposable branch; require git review, protected branches, and no production credentials.
Claude CodeCoding agentsAI-assisted launch evidence recordConditional pilotshell command executionTerminal command and file-editing power is useful for bounded repo work but requires shell, branch, secret, and review controls before pilot.healthycomments on issues or PRs, accesses secrets or environment variablesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Pilot Claude Code only in non-production repositories with branch isolation, protected branches, command restrictions, audit logging, test credentials, and mandatory human review before merge.
Cursor Background AgentsIDE agentsAI-assisted launch evidence recordConditional pilotpull-request creationBackground coding work can be piloted after remote-environment isolation and branch-review workflow are verified.healthyruns commands, opens pull requestsleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Use isolated background tasks in non-production repos; review generated branches/PRs and verify remote environment, secrets, and audit logs.
Gemini CLITerminal agentsAI-assisted launch evidence recordConditional pilotshell command executionCLI coding assistance with local command context requires terminal sandbox, command allowlist, and test credentials before pilot.healthyedits files, opens pull requestsleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Pilot Gemini CLI locally or in disposable devcontainers; restrict command execution and require human review of all diffs.
GitHub Copilot coding agentGitHub/PR agentsAI-assisted launch evidence recordConditional pilotpull-request creationGitHub-native issue-to-PR workflow is pilotable only with repository installation scope, branch protection, and PR approval rules.healthyreads repository, edits filesleast-privilege permissions, protected branches, mandatory human reviewnot reviewed2026-07-15Check
Assign the coding agent only to selected repos and non-critical issues; require protected branches, CODEOWNERS/human approval, and audit review.
Google JulesCoding agentsAI-assisted launch evidence recordConditional pilotpull-request creationAsync GitHub coding tasks are pilotable only after repo-scope and branch/PR review boundaries are explicit.healthyreads repository, runs commandsleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Use Jules on selected GitHub repos and low-risk bugs; require scoped installation, branch protection, CI, and human review.
OpenAI CodexCoding agentsAI-assisted launch evidence recordConditional pilotshell command executionCloud coding and PR handoff can be piloted after repo scope, branch policy, test credentials, and review ownership are set.healthycomments on issues or PRs, makes network/API callsleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Pilot Codex on low-risk issues in non-production repos; require scoped GitHub access, protected branches, CI, and human review before merge.
Windsurf CascadeIDE agentsAI-assisted launch evidence recordConditional pilotshell command executionIDE command/file-editing workflows can be piloted with repo scope, command restrictions, and reviewed branches.healthyreads repository, runs commandsleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Pilot Cascade on non-production repos with protected branches, command review, audit logging where available, and no production secrets.
ClineIDE agentsAI-assisted launch evidence recordSecurity reviewshell command executionIDE agent can edit files, run commands, and use browser/computer actions, so browser/session and shell controls need review.healthyreads repository, opens pull requestsleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Security-review Cline before pilot; if approved, use a test repo, devcontainer, command allowlist, session recording for browser use, and mandatory human review.
DevinCoding agentsAI-assisted launch evidence recordSecurity reviewfile modificationAutonomous software-engineering scope is high power and should enter formal security review before operational access.healthyopens pull requests, comments on issues or PRsleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Start with a security review of repo scope, execution environment, credentials, PR workflow, and auditability; only then run a bounded pilot.
Amazon Q Developer AgentCoding agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdfile modificationTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Terminal and coding agent: review official evidence before any pilot.
Bito AI Code Review AgentGitHub/PR agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdpull-request creationTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, protected branches, mandatory human reviewnot reviewed2026-07-15Check
Pull-request and code-review agent: review official evidence before any pilot.
CodeRabbitGitHub/PR agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdpull-request creationTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, protected branches, mandatory human reviewnot reviewed2026-07-15Check
Pull-request and code-review agent: review official evidence before any pilot.
ContinueIDE agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdfile modificationTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
IDE/background coding agent: review official evidence before any pilot.
Factory DroidCoding agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdfile modificationTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Terminal and coding agent: review official evidence before any pilot.
JetBrains JunieIDE agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdfile modificationTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
IDE/background coding agent: review official evidence before any pilot.
Linear AgentsGitHub/PR agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdpull-request creationTracked record has insufficient claim-level editorial evidence for rollout use.docs riskreads repository, edits filesleast-privilege permissions, protected branches, mandatory human reviewnot reviewed2026-07-15Check
Pull-request and code-review agent: review official evidence before any pilot.
OpenCodeTerminal agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdshell command executionTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Terminal and coding agent: review official evidence before any pilot.
Qodo MergeGitHub/PR agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdpull-request creationTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, protected branches, mandatory human reviewnot reviewed2026-07-15Check
Pull-request and code-review agent: review official evidence before any pilot.
Roo CodeIDE agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdshell command executionTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
IDE/background coding agent: review official evidence before any pilot.
Sourcegraph CodyIDE agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdfile modificationTracked record has insufficient claim-level editorial evidence for rollout use.repo metadata riskreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
IDE/background coding agent: review official evidence before any pilot.
SweepGitHub/PR agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdpull-request creationTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, protected branches, mandatory human reviewnot reviewed2026-07-15Check
Pull-request and code-review agent: review official evidence before any pilot.
Tabnine AI agentsIDE agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdfile modificationTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
IDE/background coding agent: review official evidence before any pilot.
Adept ACTBrowser/computer-use agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdbrowser or authenticated-session actionsTracked record has insufficient claim-level editorial evidence for rollout use.docs riskreads repository, edits filesleast-privilege permissions, sandbox, session recordingnot reviewed2026-07-15Check
Browser and computer-use agent: review official evidence before any pilot.
AirtopBrowser/computer-use agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdbrowser or authenticated-session actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, sandbox, session recordingnot reviewed2026-07-15Check
Browser and computer-use agent: review official evidence before any pilot.
Anthropic Computer UseBrowser/computer-use agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdbrowser or authenticated-session actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, sandbox, session recordingnot reviewed2026-07-15Check
Browser and computer-use agent: review official evidence before any pilot.
Asana AI TeammatesWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.docs riskreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
Atlassian RovoWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
Base44App buildersPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, separate test credentials, deployment approvalnot reviewed2026-07-15Check
App builder and deployment agent: review official evidence before any pilot.
Bolt.newApp buildersPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, separate test credentials, deployment approvalnot reviewed2026-07-15Check
App builder and deployment agent: review official evidence before any pilot.
ClickUp BrainWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
ComposioMCP/agent marketplacesPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Terminal and coding agent: review official evidence before any pilot.
DustEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
Firebase StudioApp buildersPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, separate test credentials, deployment approvalnot reviewed2026-07-15Check
App builder and deployment agent: review official evidence before any pilot.
Framer AIApp buildersPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, separate test credentials, deployment approvalnot reviewed2026-07-15Check
App builder and deployment agent: review official evidence before any pilot.
Google Vertex AI Agent BuilderEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
Gumloop AgentsWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
IBM watsonx OrchestrateEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
KiroApp buildersPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessSpec-driven app/coding workflow can be piloted in staging only with environment separation and approval controls.healthyreads repository, edits filesleast-privilege permissions, separate test credentials, deployment approvalnot reviewed2026-07-15Check
Pilot Kiro on a small non-production app; require environment separation, protected repo branches, secret isolation, and reviewer signoff.
LindyWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
LovableApp buildersPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, separate test credentials, deployment approvalnot reviewed2026-07-15Check
App builder and deployment agent: review official evidence before any pilot.
Make AI AgentsWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessTracked record has insufficient claim-level editorial evidence for rollout use.docs riskreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
ManusBrowser/computer-use agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdbrowser or authenticated-session actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, sandbox, session recordingnot reviewed2026-07-15Check
Browser and computer-use agent: review official evidence before any pilot.
MCP RegistryMCP/agent marketplacesPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Terminal and coding agent: review official evidence before any pilot.
Microsoft Copilot Studio agentsEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsEnterprise agent-platform risk depends on connectors, tenant policy, data sources, and workflow actions configured by the organization.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Review a specific agent design in a test tenant with connector scopes, DLP/policy controls, approvals, and audit logging before pilot.
MultiOn Agent APIBrowser/computer-use agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdbrowser or authenticated-session actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, sandbox, session recordingnot reviewed2026-07-15Check
Browser and computer-use agent: review official evidence before any pilot.
n8n AI AgentWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
Notion AI agentsWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.docs riskreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
OpenAI OperatorBrowser/computer-use agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdbrowser or authenticated-session actionsTracked record has insufficient claim-level editorial evidence for rollout use.docs riskreads repository, edits filesleast-privilege permissions, sandbox, session recordingnot reviewed2026-07-15Check
Browser and computer-use agent: review official evidence before any pilot.
Oracle AI Agent StudioEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.source riskreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
Pipedream AIWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
Replit AgentApp buildersPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessApp-building and deployment capabilities are high power because they can create code and may reach hosting, secrets, databases, or deployment surfaces.healthyreads repository, edits filesleast-privilege permissions, separate test credentials, deployment approvalnot reviewed2026-07-15Check
Security-review before pilot; use a test workspace, no production data, separate credentials, deployment approval, rollback, and audit trail.
Salesforce AgentforceEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsCRM/workflow agents depend on Salesforce org configuration, data access, actions, and approval flows.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Pilot in a sandbox Salesforce org with scoped data, approval flows, audit logs, and rollback plan before touching production CRM workflows.
ServiceNow AI AgentsEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.source riskreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
Slack AgentforceWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.source riskreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
SmitheryMCP/agent marketplacesPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holdworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.source riskreads repository, edits filesleast-privilege permissions, isolated branch, protected branchesnot reviewed2026-07-15Check
Terminal and coding agent: review official evidence before any pilot.
v0App buildersPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, separate test credentials, deployment approvalnot reviewed2026-07-15Check
App builder and deployment agent: review official evidence before any pilot.
WordwareWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
Zapier AgentsWorkflow agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Evidence holddeployment accessTracked record has insufficient claim-level editorial evidence for rollout use.docs riskreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Workflow and API agent: review official evidence before any pilot.
AutoGenEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Implementation-dependentworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
Browserbase StagehandBrowser/computer-use agentsPreliminary monitored record. Not yet launch-audited for a rollout decision.Implementation-dependentbrowser or authenticated-session actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, sandbox, session recordingnot reviewed2026-07-15Check
Browser and computer-use agent: review official evidence before any pilot.
CrewAIEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Implementation-dependentworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
DifyEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Implementation-dependentworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
FlowiseEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Implementation-dependentworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
Hugging Face smolagentsEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Implementation-dependentworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
LangGraphEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Implementation-dependentworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
Mistral Agents APIEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Implementation-dependentworkflow/API actionsTracked record has insufficient claim-level editorial evidence for rollout use.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Agent SDK/framework or enterprise platform: review official evidence before any pilot.
OpenAI Agents SDKEnterprise agent platformsPreliminary monitored record. Not yet launch-audited for a rollout decision.Implementation-dependentworkflow/API actionsThe SDK is not itself a deployed agent; rollout risk depends on the application, tools, data, and approval logic built with it.healthyreads repository, edits filesleast-privilege permissions, API scope restrictions, audit logsnot reviewed2026-07-15Check
Run architecture review before implementation: define tools, data stores, API scopes, human approvals, audit logs, retention, and kill-switches.