Agent Roles
Every built-in agent role in Cerevisor, what it's for, its smart defaults, and when to use which.
Cerevisor ships with 16 built-in roles plus Custom. A role is mostly a starting point; it sets a few smart defaults (which subagent type to use, which tag color the badge has), then you write the instructions that turn the role into a specific agent for your specific workflow. Roles deliberately have no fixed default model — every new agent starts on auto, and model choice is resolved at run time (see below).
You can also pick Custom and define everything from scratch, or create your own reusable agents via Settings → Agents → Custom Agents.
The full list
| Role | Tagline | Default subagent | When to pick it |
|---|---|---|---|
| Researcher | Investigate & discover | explore | Web research, source gathering, market scans, anything that involves crawling and synthesizing external information. |
| Architect | Design & plan | plan | High-level design decisions, sequencing, breaking a big goal into ordered steps. |
| Developer | Build & implement | general-purpose | Writing code, generating files, executing concrete build tasks. |
| Reviewer | Review & validate | explore | Checking another agent's output, looking for bugs, validating against criteria. |
| Documenter | Document & explain | general-purpose | Producing user docs, READMEs, API references, internal documentation. |
| Tester | Test & verify | general-purpose | Writing or running tests, verifying invariants, integration testing. |
| PowerPoint Presentation Builder | Slide decks & narratives | (auto) | Producing slide decks from research or planning output. Pairs well with the pptx skill. |
| Blog Post Writer | Long-form article drafting | (auto) | Drafting 600-2000 word blog articles. Pairs well with writing skills. |
| Excel Expert | Spreadsheet modeling expert | (auto) | Building or editing .xlsx financial models, data sheets, dashboards. Pairs with the xlsx skill. |
| Newsletter Writer | Audience-ready editions | (auto) | Putting together a recurring newsletter from upstream content. |
| SEO Content Optimizer | Search visibility tuning | (auto) | Editing existing content for search visibility — title tags, headings, keyword fit, snippet structure. |
| Meeting Notes Summarizer | Decisions & action items | (auto) | Turning raw meeting transcripts into structured summaries with decisions and action items. |
| Research Brief Writer | Structured research briefs | (auto) | Converting research findings into structured, decision-ready briefs. |
| Social Media Caption Writer | Caption & hook generation | (auto) | Producing LinkedIn / X / Bluesky captions from longer-form content. |
| Coach | Integrate & guide | general-purpose | A senior agent that reviews other agents' decisions, mentors the workflow, makes judgment calls. |
| Situation Reporter | Summarize handoff state | plan | Advisory label: shapes the agent toward composing a structured handoff (default output file: wave-situation.md) as a checkpoint between waves, so downstream agents inherit one canonical read of what's done, what's missing, and what's still unknown. The harness's own wave situation reports are role-agnostic and fire regardless. Pairs with the Situational Awareness pillar. |
| Custom | Define your own | auto | Everything starts blank. You set role description, instructions, model, tools. |
"Auto" means what?
For subagent type, auto means: at run time, the orchestrator picks a sensible default based on the agent's instructions. For the model, auto means the agent runs on its provider's configured default model, unless you've turned on smart routing (Settings → Integrations → Smart Routing) or Cerevisor has learned a better fit from your past runs. You can always override per agent.
Subagent types
The four subagent types influence how the harness scopes the agent's permissions and which tools it gets by default.
| Subagent type | Tool defaults | Best for |
|---|---|---|
| explore | Read-only tools (read, glob, grep, list_directory, find_skills, invoke_skill, todo). No write or execute. | Researchers, reviewers, agents that should look at things but not modify the world. |
| plan | Read tools + thinking modes that produce structured plans. | Architects, planning agents. |
| bash | Bash execution + file read. No file write. | Agents that run shell commands but shouldn't write files directly. |
| general-purpose | Full toolset (read, write, edit, bash, web). | Developers, testers, coaches — agents that actually do things. |
You can override the default tool set per agent in the Agent Config popup → Allowed tools.
Model preferences
Every new agent starts with its model preference on auto. You can set a tier preference per agent via the Model picker in Agent Config:
| Preference | What it means |
|---|---|
| haiku | Prefer the fast, inexpensive tier on the resolved provider. |
| sonnet | Prefer the mid tier. |
| opus | Prefer the most capable tier on the resolved provider. |
| auto | The provider's configured default model — or, with smart routing on, the router's per-agent call. |
| custom | Pin an exact model id. |
Roles never force a model. Cost/quality routing is dynamic: smart routing (Settings → Integrations) and Cerevisor's learned routing from your own run history do the matching.
Creating a Custom Agent
Two ways:
- Inline: pick Custom as the role when adding the agent. Configure everything in the Agent Config popup.
- Saved template: open Settings → Agents → Custom Agents → New. Fill out the template (name, role description, default skills, default model). Save. The template now appears in the agent palette and you can drag it onto any canvas.
Saved custom agent templates sync across all your Cerevisor workflows.
Reordering and hiding roles
Don't use the PowerPoint role? Hide it.
Settings → Agents. Drag to reorder the palette. Toggle visibility per role. Reordering affects the Shift+1–9 / Shift+0 placement shortcuts (Shift+1 = first visible role, Shift+2 = second, etc.).