Parallel Task for AI agents
Parallel Task lets agents launch deep-research and enrichment jobs, track their progress, and pull back cited results. Researchers use it to automate longer research workflows with sources attached. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.
What agents do in Parallel Task.
Start a deep research task
Launch a deep-research job on a topic and receive a cited, structured answer.
Check a task's status read-only
Poll a running research task's status before deciding whether to wait or resubmit.
Retrieve results as markdown
Pull a completed research task's result formatted as markdown for a report.
4 Parallel Task actions, graded by risk.
Every Parallel Task action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
2Look things up. Allowed by default.
- parallel_task.get_statusGetstatus
- parallel_task.get_result_markdownGetresultmarkdown
write
2Create and change things. Allow, or ask the user first.
- parallel_task.create_task_groupCreatetaskgroup
- parallel_task.create_deep_researchCreatedeepresearch
destructive
0Delete, cancel or archive. Ask first, or deny outright.
- No destructive actions.
Parallel Task in three steps.
- 01Your users connect Parallel TaskThey sign in to Parallel Task on Arc0 Connect, under your brand, and approve the access you ask for.
- 02You set the rulesReads run, and writes like “createtaskgroup” can wait for the user to approve.
- 03Any agent can actYour agent calls Parallel Task through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('parallel_task', { read: 'allow', write: 'ask', // create_task_group destructive: 'deny', }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Parallel Task connects.
Parallel Task runs its own MCP server behind OAuth. Arc0 registers the client, users approve access on Arc0 Connect, and your agent reaches Parallel Task through the same endpoint, policies and audit log as every other app.
The same Parallel Task connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →
- AUTH
- OAuth 2.0 (MCP)
- CREDENTIALS
- Per-tenant encrypted vault
- MODEL SEES
- Results only, never credentials
- AUDIT LOG
- Every call, on every plan
Use Parallel Task from any agent.
More web data and scraping apps.
Parallel Task and Arc0, answered.
Can I use Parallel Task with Claude, ChatGPT or Cursor?
Yes. Connect Parallel Task to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Parallel Task actions you allow.
How do users connect Parallel Task?
Parallel Task runs its own MCP server behind OAuth. Arc0 registers the client, users approve access on Arc0 Connect, and your agent reaches Parallel Task through the same endpoint, policies and audit log as every other app.
Which Parallel Task actions can my agent take?
4 in total: 2 read, 2 write and 0 destructive, such as “createtaskgroup”. Your policies decide which of them each agent may call.
Can I make my agent read-only in Parallel Task?
Yes. Allow read actions and deny writes in the Parallel Task policy. Your agent can still look things up, and any write it attempts is blocked and logged.
Can my own backend call Parallel Task too?
Yes. The same Parallel Task connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Parallel Task into your agent.
Your users connect Parallel Task once, under your brand. Your agent gets 4 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan