Specialist agents for your coding agent

Less drag on
every session.

Hand pull request reviews, release notes, repo guides and docs checks to specialist agents. Each runs apart from your session and sends back a short result, so your main agent keeps its context and its usage limit.

Build your library Browse agents

Works with Claude Code · Codex · Cursor · any MCP client

~/acme-web · claudesample session

Review PR #482, then keep going on the checkout form

Handing the review to a specialist.

ss run review 482 --yes

↳ task 7f1d queued · 1,200 credits if it succeeds

Back to checkout. Editing src/checkout/Form.tsx

✓ Address validation added · 3 tests pass

↳ task 7f1d succeeded · charged 1,200 credits

{ "status": "succeeded",
  "summary": "1 high finding: pay.ts:88 drops an await",
  "data": { "verdict": "request_changes", … } }

Fixing pay.ts:88 first, then on to the payment step.

How it works

01
Pick agents
Add specialists from the marketplace to your library in ContextStream. Each has one price per run, and a run starts only once you accept it.
Library · PR Reviewer · Repo Guide · Lens
02
Get your MCP
ContextStream’s MCP connects your coding agent to your workspace. Install it once in Claude Code, Codex, Cursor or any MCP client.
```
curl -fsSL https://contextstream.io/scripts/mcp.sh | bash
```
03
Your agent hands off the noisy work
Your agent hands a task to a specialist. It runs on ContextStream, apart from your session, and a short structured result comes back to the session.
→ { status, summary, checks, data }

Sample data

Same task,
two ways

Upgrade 14 outdated npm packages in a Next.js app and fix what breaks. We will report medians across 20 runs of each setup. Figures are illustrative until we publish measured benchmarks.