18
FLEET STATUS · 5/6 ONLINE
Agent fleet.
Six long-running agents share a memory store and a tool registry. Each one owns a slice of recurring work. Researcher hunts, editor cleans, planner sequences, critic stress-tests, librarian keeps the memory tidy, forecaster watches the metrics.
11
2.1
168 ms
2
188
// active agents
Six workers, six remits
agt://researcher
Researcher
Hunts source material across web, papers, and the org's own docs, then produces digestible briefs with citations. Owns 80% of the inbound web_search traffic.
// last action tool_call(web_search) → 4 results
agt://editor
Editor
Takes drafts and applies the house style: tone, hierarchy, claim-checking. Flags every unsourced sentence and either patches or returns it to the researcher.
// last action rewriting section 3, clarity pass
agt://planner
Planner
Breaks an arbitrary goal into a runnable plan with explicit dependencies. Owns the kanban for the lab itself, calling in researcher and critic on demand.
// last action awaiting next dispatch
agt://critic
Critic
Tries to break whatever the rest of the fleet ships. Adversarial probes, regression replays, sanity checks. Often the difference between a good draft and a published one.
// last action probing edge cases on EX-014 outputs
agt://librarian
Librarian
Owns memory hygiene. Compacts the long-term store, deduplicates, tags new entries, evicts stale ones. The reason memory size stays bounded.
// last action compacting last 7d into a summary
agt://forecaster
Forecaster
Time-series brain. Watches the metrics and predicts trouble: thermal events, quota burn, traffic spikes. Files alerts to the on-call console.
// last action awaiting baseline refresh
// dispatch log
Agent activity, live
Every turn, tool call, memory write. Tailing the dispatch log.
Agent dispatch log
72269 [ops] Boot sequence initiated. Pid 1 = supervisord.
72270 [k8s] Reconciled 14 / 14 desired pods. Drift OK.
72271 [fs] R2 bucket `ai-uploads` mounted (4MB used / quota 5GB).
72271 [ai] Model registry warmed: Llama-3.3-70B · SDXL · Whisper-large-v3.
72272 [gpu] Probe complete · 8 GPUs available · avg 62°C · 0 throttled.
72273 [net] Edge fanout 308/308 colos. Median RTT 21ms (p95 38ms).
72274 [tls] Cert rotation lock acquired. Chain valid for 78 days.
72275 [mod] Moderation pipeline ready. Threshold 0.82 · blocklist 12.4k entries.
72275 [agt] Agent fleet check-in: 6/6 online · 0 quarantined.
72276 [rt] WebSocket upgrade pool ready. Hold capacity 1024.
72277 [sec] Turnstile binding live. Bot pass-rate target 99.2%.
72278 [ai] Llama-3.3-70B warm-up complete. First-token latency 168ms.
72279 [ops] D1 migration check: 0 pending.
72279 [k8s] HPA scaled gpu-pool · 6 → 8 replicas (q-depth above threshold).
72280 [agt] agt://researcher initialised. Tools: web_search, code_run, files.
72281 [gpu] Job queued · id 7c4a · sdxl · 4 images · est 18s.
72282 [ai] Job 7c4a complete · 17.4s wall · neurons 21,840.
72283 [mod] Reviewed 1 image · score 0.04 · pass.
// tool registry
Shared tools
web_search Authenticated DuckDuckGo wrapper with snapshot cache. 21,409/30d http_fetch HEAD-then-GET with content-type allow-list. 8,221/30d code_run Sandboxed JS/Python via Workers. 5s wall, 128MB RAM. 14,082/30d sql Read-only D1 query. Statement allow-list, EXPLAIN gated. 6,019/30d embed Workers AI bge-large-en. Batched at 64, deduped. 94,381/30d image_gen SDXL pool. 4-image grids; refiner pass at user opt-in. 2,047/30d files R2 read/write into agent-scoped bucket prefix. 12,540/30d memory.read Vector query into long-term memory store. 38,219/30d memory.write Append + tag a memory entry; librarian decides retention. 11,847/30d