Conduit MCP
- bkauto3/Conduit
Conduit is a MCP server that runs headless browser automation with SHA-256 audit trails and Ed25519-signed proof bundles for AI agents.
About
Conduit is a Model Context Protocol server that combines headless browser control with cryptographic auditability for AI agents. developers often need an agent to log into a staging site, exercise a checkout path, or capture a compliance screenshot, but plain Playwright scripts leave no standard proof that the run happened as claimed. Conduit addresses that gap by hashing audit trail material with SHA-256 and packaging Ed25519-signed proof bundles agents can hand back to you or attach to a release ticket. The server is published as conduit-browser on PyPI at version 0.2.1 using stdio transport, which plugs into Claude Code, Cursor, and other MCP clients on machines where Python tooling is already part of your stack. It is strongest when browser automation must be defensible—security reviews, regulated workflows, or investor demos—rather than for one-off local scraping without a record.
- Headless browser automation exposed as MCP tools for agents
- SHA-256 audit trails for actions and captured artifacts
- Ed25519-signed proof bundles for verifiable run records
- PyPI package conduit-browser v0.2.1 with stdio transport
- Designed for agent-driven flows that need evidence, not silent scraping
Conduit MCP by the numbers
- Data as of Aug 10, 2026 (Skillselion catalog sync)
claude mcp add conduit-browser -- uvx conduit-browserAdd your badge
Show developers this MCP server is listed on Skillselion. Paste this into your README.
| Package | conduit-browser |
|---|---|
| Transport | STDIO |
| Auth | None |
| Repository | bkauto3/Conduit ↗ |
What it does
Let agents drive a headless browser while producing SHA-256 audit trails and Ed25519-signed proof bundles you can retain after automated checks.
Who is it for?
Best when you need auditable, agent-controlled headless browsing before release or for security-sensitive checklists.
Skip if: Simple local UI tests with no proof requirements, non-technical marketing-only image tasks, or teams that forbid automated browser access to production accounts.
What you get
After you register Conduit, agent-driven browser sessions can return hash-chained audit data and Ed25519-signed proof bundles you can verify offline.
- Agent-invokable headless browser automation via MCP
- SHA-256 audit trail artifacts from documented runs
- Ed25519-signed proof bundles suitable for release or security records
By the numbers
- Server version 0.2.1
- PyPI identifier conduit-browser
- stdio transport
Recommended MCP Servers
1stDibs MCPMCP server for browsing and searching items on 1stDibs marketplace.
2Captcha MCPFull 2Captcha API (43 tools) for Claude Code: 31 captcha solvers + management + webhook.1
4fetch MCPFetch a URL and get clean Markdown with metadata.
Acrawl MCPAutonomous web crawler.11
Agentfetch MCPToken-budgeted web fetch for AI agents — auto-routes Jina, FireCrawl, Trafilatura, PDF.3AgenticTotem Web Extractor MCPAI web extraction: send URLs + a JSON Schema, get clean structured data.How it compares
Auditable headless-browser MCP server, not a generic screenshot skill or a passive monitoring SaaS.
FAQ
Who is Conduit MCP for?
Developers and small teams whose agents must automate browsers and produce verifiable SHA-256 audit trails plus Ed25519-signed proof bundles.
When should I use Conduit MCP?
Use it in Ship when you are running security or launch checks that need a headless browser and a cryptographic record of the agent’s actions.
How do I add Conduit MCP to my agent?
Install conduit-browser v0.2.1 from PyPI, configure stdio MCP in your client, and grant the server environment access needed for headless browser runs.