Conductor vs Hermes Agent: Diffs or Profile-Based Handoffs?

6 min read Comparisons

Compare Conductor coding workspaces with Hermes Agent profiles, including memory, permission proposals, recurring work and hosting limits.

Conductor and Hermes Agent ask a developer to choose between supervising coding workspaces and coordinating named agent profiles. Conductor keeps the branch, chat and review history together. Hermes Agent gives an orchestrator profile a root task and lets it assign specialist cards, inspect completion and add work.

That distinction matters when a recurring job needs several handoffs. A workspace gives you a place to inspect a change, while a profile board makes dependencies between agent roles explicit. The next question is whether those handoffs need an executable workflow definition or whether dependency-linked cards are enough.

Conductor starts with a coding workspace

Conductor's documentation puts a coding task in its own workspace, with chats and branch history. You approve a plan before implementation, review the diff and send comments back to the agent. Cloud teammates can share chats and reassign workspaces.

Its API documentation also describes an external manager agent supervising sessions and steering stalled work. That is useful when you want to automate some supervision while keeping the workspace as the coding context. Its public docs do not establish a lead assigned to a project, goal or team.

Local Codex sessions can pursue a standing objective, with status and optional token-budget usage above the chat. The published goal is narrower than an ongoing outcome with tracked business metrics. Cloud workspaces do not support those standing goals yet, so location changes the available goal behavior.

Hermes Agent assigns work to profiles

Hermes Agent's Kanban documentation describes cards with comments, dependencies and attempt history. An orchestrator profile owns the decomposed project's root task, assigns children to specialist profiles and judges the result. People can manage and comment on cards, but the documented assignee is an agent profile.

From chat, Kanban can expand an idea into a specification or dependency-linked tasks. You can disable automatic child promotion so generated work waits for manual review. A developer supervising a scout, editor and writer can inspect the dependencies before those profiles start.

The same documentation calls the board single-host by design and uses a trusted local-user model. Treat that as a deployment boundary when evaluating shared work. A profile board on your installation is a specific coordination model, and its documented ownership should not be read as interchangeable with human task assignment.

Durable instructions and learned lessons

Conductor's agent-behavior reference describes repository preferences, instruction files and skills. Those preserve editable guidance for later coding sessions. Its public docs do not describe inspectable learned memories from earlier runs.

Hermes Agent publishes persistent notes and user preferences that inform later sessions. Its memory guide describes inspecting, editing or deleting learned memory through journey commands or the desktop panel. A corrected review convention can therefore have a documented memory path.

For your trial, choose a correction that matters to repeated work, such as an accepted output example. Check where each product stores it and how you would remove a mistaken lesson. The relevant comparison is the published correction mechanism, not a general claim that either agent becomes reliable through repetition.

Approval prompts and permission proposals

Conductor lets you configure tool approvals, including pauses before local shell commands, file edits and Model Context Protocol tools. Its security and permissions reference does not describe gathering those requests into an approval queue. Its public pages also do not describe permission-change proposals derived from approval history.

Hermes Agent's security guide describes shell approval modes, allowlists and deny rules. Dangerous-command requests can reach the terminal or messaging chat. Its approvals suggest command turns previously executed and approved commands into ranked allowlist proposals for you to accept, excluding destructive command classes.

This is a concrete difference for a developer repeatedly approving routine shell commands. Hermes Agent publishes a way to propose reusable permissions from that history. You still choose the entries, and that mechanism should be evaluated as shell-permission assistance, not unlimited authority over future work.

Schedules do not settle workflow execution

Conductor Cloud routines can start agents from schedules or GitHub actions, and webhook routines handle incoming requests. Repository setup, run and archive scripts repeat parts of the coding-project lifecycle. A webhook routine opens a fresh workspace using a saved prompt, which suits repeated independent coding requests.

Hermes Agent's cron guide describes scheduled jobs with their own instructions, skills and delivery settings. Webhooks can fire those jobs immediately. For self-hosted agent-backed jobs, the gateway must be running, and a scheduler tick skips them when it is down.

Its Kanban dependency chains provide ordered handoffs, but workflow-template routing fields are reserved for a future version and ignored by the documented kernel. Do not plan a recurring process around those fields executing today. Hermes Agent separately publishes goal completion contracts, shell quality gates and independent review for checking results.

Conductor and Hermes Agent, point by point

The table compares the same questions for each product, with a source and check date on every answer. It distinguishes workspace lifecycle scripts from profile dependencies and other narrower forms of coordination.

Both columns answer the same questions from each company's public pages. Yes means the company publishes the capability the question asks about, Partly means something narrower, No means it says it does not offer it, and Not published means its public pages do not describe it. Every answer links to its source and shows when we last checked it. The oldest check on this page was November 13, 2026.

Conductor and Hermes Agent compared, with the source and check date of each answer
What you compare Conductor Hermes Agent
Goals Conductor Partly

Local Codex sessions can pursue a standing objective, with status and optional token-budget usage shown above the chat. Its docs do not describe tracking outcome metrics or targets against that objective.

conductor.build/…/agent-modes
Hermes Agent Partly

Persistent goals keep an agent working toward an objective, with a judge and optional completion tests. Its docs describe no linked metric tracking against an ongoing business goal.

Leads Conductor Partly

The API cookbook gives an external agent a Conductor manager role to supervise coding sessions and steer stalled work. Its docs do not establish a lead assigned to a project, goal or team.

conductor.build/docs/api
Hermes Agent Yes

You can name an orchestrator profile to own a decomposed project's root task, assign child tasks to specialists, judge completion and add more work when needed.

Tasks Conductor Partly

Each coding task gets a workspace with chats, branch and review history, and Cloud workspaces can be reassigned to teammates. Its docs do not establish assigning workspace ownership to an agent.

Hermes Agent Partly

Kanban cards keep an assigned agent profile, comments, dependencies and attempt history. People can comment and manage cards, but the documented assignee is a profile name rather than a person or an agent.

Recurring work Conductor Yes

Cloud routines run agents on schedules or GitHub actions, and webhook routines run a saved prompt in a fresh workspace for each incoming request.

Hermes Agent Yes

Cron jobs run once or recur on a schedule, and webhook events can fire the same jobs immediately. Each job can carry its own instructions, skills and delivery settings.

Workflows Conductor Yes

Repository setup scripts repeat sequences such as installing dependencies, copying environment files and building each new workspace. Named run actions and archive scripts cover the rest of the coding-project lifecycle.

conductor.build/…/scripts
Hermes Agent Partly

Kanban dependency chains coordinate roles such as scout, editor and writer. They provide ordered handoffs, but the kernel does not execute workflow templates, whose routing fields are reserved for a future version.

Autonomy Conductor Partly

Tool approvals are configurable user settings, and local tool calls can pause before shell commands, file changes or MCP tools run. Its docs do not describe collecting those requests into an approval queue.

Hermes Agent Yes

You configure shell approval modes, permanent allowlists and deny rules. Dangerous-command requests appear in the terminal or messaging chat, where you can approve or deny them.

Learned autonomy Conductor Not published

Its docs describe configurable approvals and durable repository instructions, but no proposals to change an agent's permissions based on your approval history.

Hermes Agent Yes

The approvals suggest command mines executed, approved commands into ranked allowlist proposals. You choose which entries to apply, and destructive command classes are excluded.

Memory Conductor Partly

Repository preferences, instruction files and skills preserve editable guidance for later sessions. Its docs do not describe inspectable memories learned from previous runs.

conductor.build/…/agent-behavior
Hermes Agent Yes

Persistent notes and user preferences carry into later sessions. You can inspect, edit or delete learned memory through the journey commands or the desktop panel.

Planning Conductor Yes

Claude Code and Codex Plan Mode produce a plan before editing files. You can approve the approach or give feedback before moving into implementation.

Hermes Agent Yes

From chat, Kanban can expand a rough idea into a specification or dependency-linked tasks. You can disable automatic child promotion so the generated tasks wait for manual review before execution.

Intake Conductor Yes

The HTTP API creates workspaces and sends prompts to coding agents. A webhook routine binds incoming JSON to a saved prompt, project and selected agent.

conductor.build/docs/api
Hermes Agent Yes

Signed webhooks turn outside events into agent prompts. A route can target a named profile when profile multiplexing is enabled, and filter which events start work.

Connectors Conductor Yes

Agents use custom local or remote MCP servers, while the HTTP API supports workspace automation and webhook-triggered routines. MCP configuration follows the selected coding agent.

Hermes Agent Yes

A curated Model Context Protocol catalog and custom local or remote servers connect external tools. HTTP webhook routes accept signed service events and can post responses back through integrations such as GitHub.

Models Conductor Partly

You choose a model from Claude Code, Codex, Cursor or OpenCode, and the API recommends passing an explicit model. Its docs do not describe automatic model selection per task or step.

Hermes Agent Partly

You choose the main model and can override auxiliary tasks. Provider routing chooses a backend for a selected model, rather than selecting a model from each task's needs.

Cloud or your own computer Conductor Yes

Local sessions run on your Mac, while Cloud sessions run in managed Vercel sandboxes and continue with the app closed. Customer-owned cloud machines are not supported yet.

conductor.build/…/faq
Hermes Agent Yes

Run the free agent locally through its desktop or terminal interfaces, or choose the optional Hermes Cloud service for a hosted instance with a persistent workspace.

Your own subscription Conductor Yes

Claude and Codex subscriptions or provider API keys cover model usage, including in Cloud workspaces. Conductor says it does not bill or resell model usage.

Hermes Agent Yes

Hermes accepts your provider API keys and a ChatGPT account login. Its Claude account path requires Max plus purchased extra-usage credits, not the included Max allowance.

Pricing Conductor Not published

The general FAQ calls Conductor free, but the Cloud FAQ names Pro, Teams and Enterprise without amounts. It describes usage-based compute pricing as future, so current paid rates and billing basis are not published.

Hermes Agent Yes

The MIT agent is free. Portal lists Plus $20/month, Super $100/month and Ultra $200/month, with credit top-ups. Cloud's FAQ separately meters compute and storage, excluding inference and tools, without explaining how that relates to included plan hosting.

Correctness Conductor Yes

Workspace actions run tests, and Checks gathers CI status and unresolved work before merge. The API cookbook also instructs a supervising agent to review results against an implementation brief and request fixes.

Hermes Agent Yes

Goal completion contracts define verification criteria, and shell quality gates must pass before a goal completes. Kanban review also maps acceptance criteria to evidence and can return work for correction.

For software work

Conductor and Hermes Agent compared, with the source and check date of each answer
What you compare Conductor Hermes Agent
Code review rules Conductor Yes

Repository-specific Code review preferences guide the Review action, which asks an agent to inspect the current diff. You write the guidance in Repository Settings.

Hermes Agent Yes

The pull-request review guide uses reusable review guidelines and remembered team conventions, such as required type annotations and database access rules, to check later diffs.

Adversarial review Conductor Yes

The homepage demonstrates asking a different model to review a proposed fix for risks, then opening another session to challenge its approach. This is user-directed second review, not an automatic risk-triggered gate.

Hermes Agent Yes

A bundled pre-commit skill delegates security and logic review to a separate, diff-only reviewer in interactive sessions. Kanban's review skill independently verifies handoffs and keeps the reviewer from editing the implementation.

Risk rating Conductor Not published

Checks gathers merge-readiness signals such as CI, comments and todos. Its docs do not describe a risk rating of the change before merge.

Hermes Agent Not published

Its docs describe no change-level risk rating before merge. Optional residual-risk notes on Kanban handoffs and severity labels for individual review findings do not establish such a rating.

Pull requests Conductor Yes

Agents draft pull requests, receive inline review comments, fix failing checks and prepare branches for merge. Checks and the Diff Viewer keep GitHub review feedback with the workspace.

Hermes Agent Partly

Agents can open pull requests through the GitHub CLI, and the bundled GitHub skill requires reading PR comments. The cited docs do not establish a review-feedback revision cycle.

A fact we cannot find or cannot check is left out, and prices can change after the date shown. Task Machine publishes this comparison and is an interested party. Conductor and Hermes Agent are named only to identify them, and there is no relationship between the companies. Has something changed? Tell us.

Who each one suits

Pick Conductor if you want plans, diffs, tests and pull-request feedback beside coding-agent sessions. Local sessions end when the app closes or your Mac shuts down, while Cloud sessions continue with the app closed. Its Cloud FAQ names paid plans without publishing amounts, so confirm your intended Cloud terms.

Pick Hermes Agent if editable learned memory, specialist profiles and approval-history suggestions fit your operating style. The core agent is free under MIT, while Nous Portal lists Plus $20/month, Super $100/month and Ultra $200/month with credit top-ups. Its Cloud FAQ separately meters compute and storage without explaining the relationship to included plan hosting, so the self-hosted and hosted choices need separate billing checks.

Where Task Machine fits

Task Machine fits when handoffs include people owning work as well as agents. Its Workflow builder defines ordered steps, branches, human questions, approvals and verifier checks. It also proposes autonomy changes from approval and correction evidence, with your acceptance required, but is the wrong pick if direct workspace supervision or a self-hosted profile board already covers your work.

Read Task Machine vs Conductor and Task Machine vs Hermes Agent for the ownership and control differences. Before changing tools, map one repeated job into its handoffs and mark which must wait for a person, then compare that map with the workflow rows on those pages.