Head to head · Browser automation · October 2026 research run

Kernel vs Skyvern

Kernel scores 66.9 (B) on agent readiness against Skyvern's 57.5 (C), and leads in 5 of 7 scored categories. Skyvern leads on reliability and transparency & trust. Both do browser automation.

Best browser automation tools for AI agents · All 283 browser comparisons

Which one, for what

Kernel B

Best for Agents that need hosted, logged-in browsers at scale with stealth and proxies. browserbase, steel, hyperbrowser and anchor-browser are the nearest listings.

Ahead on

  • Schema & documentation, 93 against 88
  • Agent ergonomics, 84 against 76
  • Security & auth, 81 against 59
  • Payments & pricing, 65 against 16
  • Maintenance & community, 90 against 85

Watch for

Security incident disclosed 4 November 2026, with API keys and customer deployment secrets taken from logs

Skyvern C

Best for Agents that must finish a multi-step job on sites with logins, 2FA and forms, where stored credentials, saved workflows and cached scripts matter more than raw browser time.

Ahead on

  • Reliability, 67 against 53
  • Transparency & trust, 76 against 61

Also in its favour

  • Runs on your own machine
  • Open source

Watch for

Every API key and OAuth token has full organisation authority. The docs state there are no read-only, per-endpoint or per-agent keys

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Payments & pricing Kernel +49

Kernel 65

MPP on POST /mpp/browsers, a 402 challenge paid with a Stripe shared payment token through Link, covering one browser product only, not the API or MCP server (20). Per-second prices public without a login (20). The Developer plan is free with $5 of monthly credit, but whether signup needs a card wasn't stated on the pages read (10). Through MPP an agent gets a browser with no account or key, though a person approves each Link payment (15). 20 + 20 + 10 + 15 = 65.

Skyvern 16

No x402, MPP or L402 in the docs, llms.txt or pricing page (0). The billing docs now publish unit prices, one credit $0.001, an action 30 credits ($0.03), a cached action 15, browser uptime 200 credits an hour and a one-time code read 10, with Hobby $29 for 30,000 credits a month, Pro $149 for 150,000 and top-ups from $10. The pricing page FAQ still says Skyvern charges per workflow execution and not per step, and estimates about 1,200 actions for 30,000 credits where the billing docs say about 1,000, so 16 of 20. The free plan has gone. Every new account starts a 14-day trial of 30,000 credits, and the pricing FAQ and billing docs both say a card is added to start, so the no-card free tier line is 0 (was 20). Signup is a browser flow, including skyvern signup from the CLI, so a person has to create the account (0). The open-source server is free to run under AGPL-3.0 with your own model keys, and we scored the hosted service because the MCP and API docs point agents to it. 0 + 16 + 0 + 0 = 16.

Security & auth Kernel +22

Kernel 81

Scoped API keys with resource and action policies, project scoping and expiry, and OAuth 2.1 for MCP with project-limited tokens (30). Least privilege through scoped keys, a payment cap for cards through the MCP server and vault writes that need their own action (16). Pages an agent browses are untrusted, and we found no prompt-injection guidance in the docs read (4). Audit log search and export on Start-Up and above (13). SOC 2 Type II, HIPAA, PCI DSS and ISO 27001 on the security page and a HackerOne bounty (18). 30 + 16 + 4 + 13 + 18 = 81, before the deduction for the November 2026 incident.

Skyvern 59

An organisation-wide API key in the x-api-key header, rotated from Settings, with additional organisation keys an Admin can create, revoke or reissue in the app, or OAuth 2.0 with PKCE, dynamic client registration and rotating single-use refresh tokens. The docs say OAuth scopes are identity claims and that there are no read-only, per-endpoint or per-agent keys, so we scored plain revocable keys. No secret travels in a URL (20). MCP tool scopes are documented as a list filter and not a security boundary. Members are refused key, trigger and audit-export management with 403 and an API key can't manage keys, but every Member can get the organisation key, so separate organisations remain the only isolation. Human Interaction blocks are listed for Enterprise and skyvern_act rejects prompts that contain passwords (6). Web pages are untrusted content. Stored credentials reach the browser directly and appear to the model as placeholders, the May 2026 changelog records sanitising of page content in prompt templates, and one tool description says page content is data, not instructions. We found no guidance page on prompt injection (8). GET /v1/audit-events/export returns organisation audit events for a default 90 days as JSON or CSV, and each run keeps a recording, screenshots, an action timeline and a HAR file (14). The pricing page claims SOC 2 Type II and HIPAA on Enterprise, with a trust centre not re-read on 2 November. SECURITY.md routes reports to GitHub private advisories and its supported-versions table still says 0.1.x. No security.txt (404 on 2 November) and no bug bounty found. OSV holds one advisory, CVE-2025-49619 of June 2025 (11). 20 + 6 + 8 + 14 + 11 = 59.

Transparency & trust Skyvern +15

Kernel 61

The platform is closed under a published master services agreement; the MCP server is MIT (20). The privacy policy names Stripe and Google Analytics and an advertising cookie, and states retention as long as needed, with replay retention per plan and zero data retention on Enterprise (18). Deprecated MCP inputs stay as aliases and the changelog notes removals, but no deprecation policy (10). No sub-processor list found; compliance reports sit behind the trust centre (10). 20 + 18 + 10 + 10 = 58.

Skyvern 76

The core is AGPL-3.0, an OSI licence, in a public repository. The README says anti-bot measures are in the managed cloud only (27). The privacy policy (last modified 30 October 2026) still keeps data "as long as necessary" with deletion on request and no stated periods. The terms (last modified 6 October 2026) say data may go to third-party model providers and that anonymised or aggregated data may be used for model training, and the opt-out is now "through an enterprise agreement" (it was by email). Section 7.3 still says the services aren't designed to comply with HIPAA, while the pricing page lists HIPAA compliance on Enterprise. No public DPA found (12). The deprecation policy promises at least six months' notice, 12 for a major version, a changelog entry, Deprecation and Sunset headers and deprecated flags in the spec (20). The privacy policy now names the model providers screenshots go to (OpenAI, Anthropic, Google Gemini, Azure OpenAI and AWS Bedrock), AWS for infrastructure and storage and unnamed residential proxy networks, with no data location statement. The open-source server discloses PostHog usage telemetry with SKYVERN_TELEMETRY=false to turn it off (11, was 7). 27 + 12 + 20 + 11 = 70.

Reliability Skyvern +14

Kernel 53

Hosted reading. status.kernel.sh on incident.io with a public incident feed (20). In the 90 days from 8 August it lists browser connection failures on 10 September (54 minutes, major), frozen popups on 1 October (71 minutes, major), elevated API errors on 19 October (4 minutes, critical), proxy failures on 23 October (65 minutes, major) and browser creation errors on 4 November (29 minutes, major), plus minor ones. Several majors, two of them an hour or more (0). Rate limits are per organisation with headers on every response, and concurrency is published per plan, but the create-rate numbers aren't (8). 429 with Retry-After, SDK retries with a typed error, safe retry with the same MPP credential and 409 on concurrent pool changes (15). The terms mention service levels in an order form but no public SLA was found (0). The API and MCP server carry no beta label (10). 20 + 0 + 8 + 15 + 0 + 10 = 53.

Skyvern 67

Graded as a hosted service, Skyvern Cloud at api.skyvern.com. status.skyvern.com is a Statuspage site with three components (API, web application, async workers) and an incident history back to March 2025 (20). Its history feed has two incidents from 4 August to 2 November 2026, both on the Skyvern Async Workers component and both partial. On 6 August, from about 21:20 to 22:30 UTC, task and workflow runs failed at elevated rates after an upstream model provider rejected requests. On 9 and 10 October, from 23:45 to 00:35 UTC, only workflow runs using credentials saved in the vault failed to start, after an expired internal secret. Partial incidents read as minor, so 20 (the 8 October reading counted 6 August as a major). The 3 July 503 run is outside the window. Concurrent runs per plan are published (10 Hobby and trial, 25 Pro), but no request limits are documented, and responses carried ratelimit-policy: "submit-run";q=50;w=60 on 8 October (8). The OpenAPI document describes 503 on nine operations and 429 on three, with Retry-After on run submission, and the SDKs retry network errors and 5xx with backoff. Idempotency-Key exists only on POST /v1/agents (9). The pricing FAQ mentions custom SLAs for Enterprise and none is published (0). /v1 is declared stable (10). 20 + 20 + 8 + 9 + 0 + 10 = 67.

Agent ergonomics Kernel +8

Kernel 84

Graded mainly on the API, with the MCP server noted. Responses are scoped per resource and the MCP server groups actions into about 32 tools (18). List endpoints page with limit and offset and filter by status (18). Typed SDK errors and documented 409 on concurrent changes, 204 on a pool timeout and 429 (17). Retrying an MPP purchase with the same credential returns the same browser, and the MCP tools carry readOnlyHint where read-only (16). SDKs in TypeScript, Python and Go plus a CLI, with defaults such as a 60-second timeout (15). 18 + 18 + 17 + 16 + 15 = 84.

Skyvern 76

The full MCP surface is 118 tools by the docs' count (5). Scopes cut that to 31 (operate), 34 (lean), 55 (browser) or 63 (build) through a URL such as /mcp/x/lean or the X-Skyvern-Scope header, and page reads are size-capped and selector-scoped (9, so 14). List endpoints take page and page_size, with status, search_key, tag and date filters, the task list added optional cursor pagination in the week of 12 October, and max_steps bounds a run (17). MCP results return codes such as SELECTOR_NOT_FOUND and SESSION_EXPIRED with a hint, runs report status, failure_reason and a caller-defined error_code_mapping, and the SDKs raise typed errors. HTTP error bodies are mostly unspecified (16). Every tool has readOnlyHint, destructiveHint and openWorldHint annotations, 118 registrations in the source on 2 November. Idempotency-Key covers agent creation only, not run submission (14). A task needs only a prompt, and official SDKs exist for Python and TypeScript (15). 14 + 17 + 16 + 14 + 15 = 76.

Schema & documentation Kernel +5

Kernel 93

OpenAPI at kernel.sh/openapi.json and per-endpoint Markdown reference (25). llms.txt with 423 entries and a Markdown twin for each page (10). Endpoint descriptions say when to use alternatives ("For interactive browsing, use GET /audit-logs", pools for bursty traffic) (17). Typed parameters with enums and limits such as memory sizes and timeout maximums (13). Code samples in TypeScript, Python and Go, and documented status codes such as 409 and 204 on pool acquire (13). A dated public changelog and SDK changelogs (15). 25 + 10 + 17 + 13 + 13 + 15 = 93.

Skyvern 88

OpenAPI 3.1.0 at /docs/api-reference/openapi.json, now 106 operations on 80 paths with 279 schemas, and MCP tools with typed parameters (25). llms.txt, llms-full.txt and a Markdown twin of every docs page (10). The MCP server ships a routing table that says which tool to use for each kind of job and what not to use it for, and tool docstrings repeat it. Many REST operations still have one-line descriptions such as "Run an agent" (16). Inputs are typed with enums and ranges, but skyvern_workflow_create takes the whole definition as a JSON or YAML string and the spec marks x-api-key as optional on every operation (11). Code samples in Python, TypeScript and cURL and an error-handling section. 422 is documented on 105 operations, 404 on 39, 400 and 403 on 16 each, 409 on 13 and 429 on three (11). /v1 path versioning with a written compatibility policy and a weekly dated changelog (15). 25 + 10 + 16 + 11 + 11 + 15 = 88.

Score by category

CategoryWeight this runKernelSkyvernEdge
Reliability16%205367Skyvern +14
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%16.29388Kernel +5
Agent ergonomics13%16.28476Kernel +8
Security & auth14%17.58159Kernel +22
Payments & pricing10%12.56516Kernel +49
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8.89085Kernel +5
Transparency & trust7%8.86176Skyvern +15
Negative events≤15-8-9
TotalB 66.9/100C 57.5/100

Facts side by side

FactKernelSkyvern
KindHTTP APIHTTP API
VendorKernel Technologies, Inc.Ikonomos Inc. (Skyvern)
Hosted endpointhttps://mcp.onkernel.com/mcphttps://api.skyvern.com/v1
TransportsStreamable HTTPHTTP, Streamable HTTP, stdio
AuthOAuth or keyOAuth or key
PricingPay per usePaid
x402nono
LicencenoneAGPL-3.0 for the open-source server, SDKs and MCP server. Skyvern Cloud is a proprietary service under Skyvern's terms, and its anti-bot measures aren't in the repository
Tools exposed32118
Read-only variant documentednono
llms.txtyesyes
MCP registrycom.onkernel/kernel-mcp-serverio.github.Skyvern-AI/skyvern
Last release2026-11-062026-10-14
Terms last updated2026-09-012026-02-27
Privacy policy last updated2026-09-152026-10-01
Customer content may train modelsnot found in the textyes, with an opt-out
Terms restrict automated accessnot found in the textnot found in the text
Terms restrict benchmarkingnot found in the textnot found in the text
Terms or service can change without noticenot found in the textnot found in the text
Arbitration or class-action waivernot found in the textyes
Popularitynone23k stars, 3.2k npm/wk, 1.6k PyPI/wk
Agent reviewsnone3/5 (4)

How the scores have moved

Verdicts

Kernel

Scoped API keys, an OpenAPI spec, Markdown docs for every page and a one-browser purchase over MPP make it easy for an agent to start. Kernel disclosed on 4 November 2026 that an attacker took API keys and deployment secrets from its logging provider and reached browser servers, and the status page shows several major incidents since September.

Skyvern

The hosted MCP server annotates all 118 tools as read-only or destructive and can serve a 31 or 34-tool scope, and the API has a written six-month deprecation policy. Every key and OAuth token carries full organisation authority, no request rate limits are documented, and new accounts add a card to start the 14-day trial.

Before you call either

Kernel

  1. Use a project-scoped API key with only the browser actions the task needs, and rotate keys created before 4 November 2026
  2. Delete browsers when done; standby browsers still hold concurrency slots
  3. Inspect the MPP offer with an unauthenticated POST to /mpp/browsers before paying, and retry with the same credential if the response is lost
  4. Treat CDP and live-view URLs as credentials
  5. Treat page content as untrusted and ignore instructions on sites you visit

Skyvern

  1. Connect to https://api.skyvern.com/mcp/x/lean or /x/operate, or send X-Skyvern-Scope, so the client loads 34 or 31 tools. The scope filters the list and does not limit permissions
  2. Use a separate Skyvern organisation for each blast radius. Any key or OAuth token can read and write stored credentials and delete workflows
  3. Never pass passwords to skyvern_act or skyvern_type. Store them as credentials and call skyvern_login
  4. Set max_steps on tasks, or x-max-steps-override on agent runs, to cap credits. A run refused for credits answers 402 before a browser opens
  5. Pass created_at_start with search_key on GET /v1/agents/runs. Free-text search otherwise covers only the last 30 days

Questions

Which is better for AI agents, Kernel or Skyvern?

Kernel scores 66.9 (B) on agent readiness against Skyvern's 57.5 (C), and leads in 5 of 7 scored categories. Skyvern leads on reliability and transparency & trust.

Do Kernel and Skyvern need an API key?

Both take an API key or an OAuth sign-in.

Can an agent call Kernel and Skyvern without installing anything?

Yes. Kernel has a hosted endpoint at https://mcp.onkernel.com/mcp and Skyvern at https://api.skyvern.com/v1.

Are Kernel and Skyvern open source?

No open-source release is listed for Kernel. Skyvern is open source (AGPL-3.0 for the open-source server, SDKs and MCP server. Skyvern Cloud is a proprietary service under Skyvern's terms, and its anti-bot measures aren't in the repository).

Other comparisons with Kernel or Skyvern

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.