feat(remote_agent): update remotebuddy implementation

[PiyushDatta]

Summary

• Apply WorkerPal completion e0821442-60bf-4921-b776-4e63cb3037f0 to main_agents.
• Integrate commit 7814d55057ee46d4e4421603e02f39c8e65db559 from refs/pushpals/agent/workerpal-6d6fd50e/e0821442-60bf-4921-b776-4e63cb3037f0.
• Worker workerpal-6d6fd50e reported: Executed task and modified 1 file(s)
• Canonical task request: Implement adjacency-aware bias in apps/remotebuddy/src/autonomous_engine.ts by constructing a queue/worker opportunity graph, scoring planner objectives with adjacency weighting under policy guardrails, and ensure reliability-heavy pro...

Motivation / Context

• Preserve and review autonomous worker output before final merge to base branch.
• Keep integration branch current with queued worker completions.

Planned Scope

• apps/remotebuddy/src/autonomous_engine.ts

Planned Validation

• Planned: bun run test:root

Changes

• Updated apps/remotebuddy/src/autonomous_engine.ts

Testing / Validation

• Planned: bun run test:root
• Worker completion summary did not include explicit command pass/fail output.

Impact / Risk

• Risk level: medium (automated worker-generated change; maintainer review required).
• No secrets or credentials are expected in this PR body.

SourceControlManager Note

• Use this worker-provided PR title/body when creating the integration PR.
• Suggested title: fix(repo): Implement adjacency-aware bias in apps/remotebuddy/src/autonomous_engine.ts by constructing...

Checklist

• Tests added/updated where appropriate

• Validation commands run (or noted as not run)

• Docs/comments updated if needed

• No sensitive data (secrets/tokens) committed

• Agent branch: agent/workerpal-6d6fd50e/e0821442-60bf-4921-b776-4e63cb3037f0

• Completion ref: refs/pushpals/agent/workerpal-6d6fd50e/e0821442-60bf-4921-b776-4e63cb3037f0

• Commit: 7814d55057ee46d4e4421603e02f39c8e65db559

• Completion ID: 678ce66d-6277-457a-89cb-a0783f82e814

[PiyushDatta]

ReviewAgent: Changes Rejected (score 7.6/10)

Verdict: The change adds meaningful scheduling intelligence, but it introduces behavior-regression risk in reliability task selection and lacks the test coverage needed for production confidence.

Issues:

• Reliability gating now hard-blocks all non-question reliability candidates when latency telemetry is missing (slackOk = params.latencyEvidenceOk && ... in evaluateReliabilityCapacityGate), which can regress existing behavior in environments without queue telemetry; add a degraded-mode path that permits bounded reliability work when evidence is absent.
• No tests are included for parseTelemetryEvidence/normalizeTelemetryValue; add unit tests for ms/s/us/% parsing, mixed separators, malformed tokens, and key collisions to prevent silent metric misinterpretation.
• No tests cover buildQueueWorkerOpportunityGraph and computeAdjacencyBias; add deterministic tests validating queue pressure/slack calculations, guardrail behavior under different policy risk/breadth settings, and adjacency boost bounds.
• No integration-level tests verify candidate filtering and scoring after the new reliability gate + adjacency boost logic; add tests asserting reliability candidate keep/drop behavior (including requires_user_input exception), and score ordering stability with/without telemetry.
• Telemetry evidence accounting can overstate confidence (latencyEvidenceCount sums hint and latency-derived counts separately), which may mark sparse telemetry as stronger than it is; track unique queue samples contributing evidence or separate counts in gating logic.

This PR has been re-queued for automated fixes. A worker will address the issues above.

[PiyushDatta]

ReviewAgent: Changes Rejected (score 7.8/10)

Verdict: Feature scope is strong, but there are correctness and regression risks that should be fixed before treating this as production-ready.

Issues:

• Potential runtime crash in telemetry identity normalization: normalizeTelemetryTimestampToken calls new Date(...).toISOString() on untrusted numeric/string metadata without guarding invalid/out-of-range timestamps, which can throw RangeError and abort candidate processing in extractQueueTelemetryFromSignals (apps/remotebuddy/src/autonomous_engine.ts).
• Behavioral regression risk: server event type changed from question_answered to generic log for autonomy question resolution payloads (apps/server/src/server_main.ts). Any downstream consumer keyed on question_answered will silently stop working; no compatibility shim or migration path is shown.
• Startup preflight interface contract expanded (readBunVersion, readDockerVersion, telemetry) but this PR does not demonstrate comprehensive callsite/test coverage for all StartupChecklistContext providers and failure modes; this is high risk for startup-time breakage or unhandled exceptions (apps/remotebuddy/src/startup/checklist.ts).
• Test coverage is too narrow for the new telemetry pipeline: only a single happy-path Ollama test is added, with no negative-path assertions for telemetry POST failure/timeouts, malformed usage payloads, auth failures (400/401), or ensuring generation continues when telemetry reporting fails (tests/remotebuddy.llm-telemetry.test.ts).

This PR has been re-queued for automated fixes. A worker will address the issues above.