Repository navigation
0.90.43: LiveRun keeps pursuit-only agents and shows unreadable roots; sandbox-ui 1.1.4 - #347
Merged
Merged
Conversation
…sandbox-ui 1.1.4 A run with no spawn journal (a retained or observer-only run) lists its agents only in TopSnapshot.pursuit.nodes. LiveRun now adds those agents as cards and takes each worker's missing start and end from the node's timing.startedAt and settledAt, so the AgentGraph spawn tree and the TraceWorkspace waterfall keep every recorded agent. sandbox-ui 1.1.4 brings the live failed-card fix (charts 0.21.21). 0.90.42 is the first release carrying the ./workspace-viewer.js export from #337, which agent-runtime 2.0.1's agent-runtime-top --web resolves by default.
… agents" Runtime 2.0.1 turns a settled supervise root into invalid-state with zero runs (nested worker runIds in observer.jsonl). LiveRun only printed diagnostics for the selected run, so with no run it claimed the root held no agents. Every partial root now states why.
drewstone
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Step 4 of epic #10550 (ops-board #10553): the live agent canvas under
agent-runtime-top --web(agent-runtime 2.0.1).Problem. #337 already mounts
AgentGraph(live, tree|hub) besideTraceWorkspace. Two gaps remained:TopSnapshot.pursuit.nodes. LiveRun ignored them, and the waterfall dropped any worker withoutstartedAt.partial/invalid-statewith no runs, LiveRun printed "No recorded agents in these roots" and hid the diagnostic.Change (
src/workspace/LiveRun.tsx, +15/−9):timing.startedAtandsettledAt../workspace-viewer.jsexport).There is no
usdKnownfield in AgentGraph or FlowSpan. Unknown spend already leavescostUsdunset (cost()) and never drops an agent.Proof. A real Runtime job:
supervisePursuit(), Runtime's durable entry. The router brain (zai/glm-5.3) spawned 3 router workers, withrunDir=<root>/<runId>, underagent-runtime-top --web --root <root> --assets <this build's dist>(Runtime 2.0.1). The runliverun-proof-5finished withwinner.invalid-statewith 0 runs, because nested worker runIds land in observer.jsonl. This PR shows that diagnostic instead of "No recorded agents". The fix is tangle-network/agent-runtime#2167, shipping in 2.0.2.The recordings (full desktop and phone webm, 7,419 and 7,788 frames) are held privately, as the README's public-repo rule requires.
Gate. beelink2 ran
pnpm build && pnpm typecheck && node --test tests/*.test.mjs(157/157)&& pnpm build:example && pnpm packand passed at a32d9b3. The merge gate runs at the head.