You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/architecture/browser-broker.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -17,7 +17,7 @@ The `@openclaw/sparsekernel-browser-broker` adapter materializes the ledger leas
17
17
18
18
When `OPENCLAW_RUNTIME_BROWSER_BROKER=cdp` and `OPENCLAW_SPARSEKERNEL_BROWSER_CDP_ENDPOINT=<loopback endpoint>` are set, the embedded OpenClaw browser tool receives an internal SparseKernel proxy instead of raw CDP access. Set `OPENCLAW_RUNTIME_BROWSER_BROKER=managed` to let SparseKernel ask the existing OpenClaw browser control service to start the managed browser and return its loopback CDP endpoint; `OPENCLAW_SPARSEKERNEL_BROWSER_CONTROL_URL` overrides the default `http://127.0.0.1:18791` control URL.
19
19
20
-
Set `OPENCLAW_RUNTIME_BROWSER_BROKER=native` to let SparseKernel launch and supervise a local Chromium-compatible process pool by trust zone and profile. The native pool uses a loopback-only remote debugging endpoint, a runtime-owned browser profile directory, and pooled process refcounts; the leased CDP context is released first, then the browser process is stopped after the pool idle timeout. `OPENCLAW_SPARSEKERNEL_BROWSER_MAX_CONTEXTS` caps simultaneous leased contexts per native pool and defaults to `8`. Use `OPENCLAW_SPARSEKERNEL_BROWSER_EXECUTABLE` when Chrome/Chromium is not discoverable on `PATH` or a common platform path. Headless mode is on by default. `OPENCLAW_SPARSEKERNEL_BROWSER_NO_SANDBOX=1` is an explicit opt-out and should only be used when the host environment cannot run Chromium's sandbox.
20
+
Set `OPENCLAW_RUNTIME_BROWSER_BROKER=native` to let SparseKernel launch and supervise a local Chromium-compatible process pool by trust zone and profile. The native pool uses a loopback-only remote debugging endpoint, a runtime-owned browser profile directory, and pooled process refcounts; the leased CDP context is released first, then the browser process is stopped after the pool idle timeout. `OPENCLAW_SPARSEKERNEL_BROWSER_MAX_CONTEXTS` caps simultaneous contexts per native process pool and defaults to `8`; the SparseKernel ledger also enforces the global `resource_budget.browser_contexts_max` lease cap, which defaults to `2` for small machines. Use `OPENCLAW_SPARSEKERNEL_BROWSER_EXECUTABLE` when Chrome/Chromium is not discoverable on `PATH` or a common platform path. Headless mode is on by default. `OPENCLAW_SPARSEKERNEL_BROWSER_NO_SANDBOX=1` is an explicit opt-out and should only be used when the host environment cannot run Chromium's sandbox.
21
21
22
22
Use `openclaw runtime network-proxy set --trust-zone <id> --proxy-ref <loopback-url>` to attach an existing proxy to a trust-zone network policy, or `openclaw runtime egress-proxy --trust-zone <id> --attach` to start the built-in policy-checking HTTP/CONNECT proxy and attach it. The Rust daemon exposes matching local API endpoints for trust-zone proxy attachment and supervised egress proxy lifecycle; it starts the built-in proxy by default and only launches an operator command when explicitly configured. Set `OPENCLAW_RUNTIME_BROWSER_REQUIRE_PROXY=1` when a trust zone must use a proxy-backed browser egress path. The trust zone's network policy must contain a loopback `proxy_ref`, and native browser pools launch Chromium with `--proxy-server=<proxy_ref>`. Static or externally managed CDP endpoints are rejected in this mode unless `OPENCLAW_RUNTIME_BROWSER_EXTERNAL_PROXY_OK=1` asserts that the external browser process is already proxy-controlled. This protects the SparseKernel-owned browser process path; it is not host-level egress enforcement for arbitrary host processes.
Copy file name to clipboardExpand all lines: docs/architecture/four-gb-vm-design.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -23,6 +23,6 @@ Five hundred logical agents are feasible because most are parked in SQLite as co
23
23
24
24
Book-writing and file-writing agents can run at higher active counts than coding agents because they do not all need browsers, sandboxes, test runners, or heavy model contexts. Expensive work should be scarce, leased, and scheduled.
25
25
26
-
Resource leases let SparseKernel answer which task owned which expensive resource and when it was released or expired. The ledger seeds small-VM resource budgets in `runtime_info` and enforces active task budgets during atomic task claiming: active steps, model calls, file patch jobs, and test jobs can be capped before a worker materializes a harness. Trust-zone budgets are enforced at lease creation for sandbox work: `max_processes` caps active sandbox leases, and `max_runtime_seconds` clamps lease runtime and expiry. This keeps a 4 GB machine from materializing more heavy execution work than its configured task and trust-zone budgets allow.
26
+
Resource leases let SparseKernel answer which task owned which expensive resource and when it was released or expired. The ledger seeds small-VM resource budgets in `runtime_info` and enforces active task budgets during atomic task claiming: active steps, model calls, file patch jobs, and test jobs can be capped before a worker materializes a harness. Browser context and heavy sandbox budgets are enforced at resource-lease creation, so a second pool or trust zone cannot bypass the global small-VM cap. Trust-zone budgets are also enforced at lease creation for sandbox work: `max_processes` caps active sandbox leases inside the zone, and `max_runtime_seconds` clamps lease runtime and expiry. Operators can tune these values with `openclaw runtime budget set --active-agent-steps-max <n> --browser-contexts-max <n> --heavy-sandboxes-max <n>`. This keeps a 4 GB machine from materializing more heavy execution work than its configured task and trust-zone budgets allow.
27
27
28
28
Browser targets and observations are compact ledger rows, not retained screenshots or traces. This lets small machines keep enough browser provenance to answer which target made a request, emitted console output, or produced an artifact while still pruning old observations with `openclaw runtime prune`.
0 commit comments