rdmbair15m5-changelog-20260923-0905-feat-20260905-23-tyrell-telemetry-and-bus-observation

One-line summary: Implemented FleetTelemetryCoordinator in TyrellCore absorbing host state, health pass, agent status, and bus observation into native Swift, resolving FEAT-20260905-23, ISSUE-20260901-27, and ISSUE-20260905-04.

1. Scope

• Hosts Affected: Fleet-wide (rdmsm4x, rdmbair13m5, rdmbair15m5, jdmbair13m5, rdmpw3265m, rdmpw3275m).
• Repositories: apps/Tyrell (SwiftPM package at git.ecs0.net:git/apps/Tyrell.git).
• Tickets: FEAT-20260905-23 (resolved), EPIC-20260905-03, ISSUE-20260901-27, ISSUE-20260905-04.

2. Exact Files Touched

• Sources/TyrellCore/Telemetry/FleetTelemetryCoordinator.swift (new, 841 lines):
  • HostHeartbeatRecord: Backward-compatible JSON schema matching agent_status.zsh and fleet_audit.zsh, including optional measured_at, enriched active_sessions, system load, memory, swap, disk, health metrics, and dirty worktrees.
  • Coupled heartbeat writer + distributor + bus sync (syncHeartbeatsAndBus) preserving Constraint 1.
  • Pure Swift bus inbox evaluator (evaluateBusWatch) actively monitoring maildir, managing ~/.agent-coordination/UNREAD-HIGH-PRIORITY.md lifecycle.
  • Host state report generator (generateStateReport) writing to ~/.agent-coordination/staging/fleetstate/<host>-<timestamp>.txt.
  • Fleet host aggregator (aggregateFleetHosts) evaluating all 6 canonical fleet hosts with status and tag filtering.
• Sources/tyrell/Tyrell.swift (modified, +149 lines):
  • Added CLI subcommands:
    • tyrell fleet hosts [--tag <tag>] [--json] [--verbose]
    • tyrell fleet telemetry [--heartbeat] [--sync] [--bus-watch] [--state-report] [--all] [--json]
  • Fixed argument parser positional subcommand handling in ECS0CommandLineParser.
• Sources/tyrelld/Daemon.swift (modified, +18 lines):
  • Added REST endpoint GET /api/fleet/hosts across both client and server data planes.
  • Added background 300s telemetry periodic task executing FleetTelemetryCoordinator.shared.syncHeartbeatsAndBus().
• Tests/TyrellCoreTests/FleetTelemetryCoordinatorTests.swift (new, 214 lines):
  • 6 unit test cases covering schema backward-compatibility, enriched heartbeats, bus watch alert file lifecycle, fleet host aggregation, and state report generation.

3. Commands Run

# Run unit tests locally
build_slot.zsh wrap swift test --package-path apps/Tyrell --filter FleetTelemetryCoordinatorTests

# Commit & push to remotes
git -C apps/Tyrell commit -m "feat(telemetry): absorb host state, health pass, agent status, and bus observation per FEAT-20260905-23"
git -C apps/Tyrell push origin main
git -C apps/Tyrell push backup main

# Fast-forward canonical on rdmsm4x
ssh rdmsm4x 'git -C ~/dev/apps/Tyrell fetch fleet && git -C ~/dev/apps/Tyrell merge --ff-only fleet/main'

# Test run on rdmsm4x
ssh rdmsm4x 'build_slot.zsh wrap swift test --package-path ~/dev/apps/Tyrell --filter FleetTelemetryCoordinatorTests'
ssh rdmsm4x 'build_slot.zsh wrap swift run --package-path ~/dev/apps/Tyrell tyrell fleet hosts'
ssh rdmsm4x 'build_slot.zsh wrap swift run --package-path ~/dev/apps/Tyrell tyrell fleet telemetry --bus-watch'

# Ticket resolution
ssh rdmsm4x 'ticket resolve FEAT-20260905-23 ...'

4. Verification Evidence

• Local Tests: Executed 6 tests, with 0 failures (0 unexpected) in 0.348 seconds.
• Remote Tests on rdmsm4x: Executed 6 tests, with 0 failures (0 unexpected) in 0.348 seconds.
• CLI Output (tyrell fleet hosts): Evaluated 6 nodes (rdmsm4x, rdmbair13m5, rdmbair15m5, jdmbair13m5, rdmpw3265m, rdmpw3275m), all reporting ONLINE with slot capacities, session counts, dirty worktrees, unread bus counts, and tags.
• Bus Watch: Evaluated maildir in <10ms, detected high-priority messages, and generated UNREAD-HIGH-PRIORITY.md.
• Commit: 3416dd2 on main.

5. Backup Locations & How to Undo

• Git Commit: 3416dd2 in apps/Tyrell.
• To Undo: git -C apps/Tyrell revert 3416dd2 git -C apps/Tyrell push origin main && git -C apps/Tyrell push backup main

6. Outstanding Owner Actions

• None. All changes are backward compatible with legacy shell cron jobs (agentstatus, buswatch, fleetstate, fleethealth).
• After a 7-day observation period showing 100% agreement between Tyrell telemetry and shell scripts, legacy launchd jobs can be demoted and archived.