| #818 |
Workflows list perf: dashboard fetches + client-groups ALL runs (9845); add grouped-summary endpoint (#agentic)
The dashboard workflows page fetches every run (days-windowed, ~9845 for a busy tenant) with no work...
|
open |
high |
2026-07-10 22:08 |
- |
|
| #817 |
DB failover breaks durability: duplicate side effects (exactly-once) + run failures under postgres restart (#agentic)
Chaos harness with --db-restarts 3 (postgres restarts mid-flight) on 20 canary runs: 3 runs FAILED a...
|
closed |
high |
2026-07-09 01:40 |
- |
|
| #816 |
Production-readiness: fault-injection / chaos harness for the durable engine (#agentic)
Add tests/chaos: a fault-injection harness that SIGKILLs workers mid-flight and asserts durability i...
|
closed |
high |
2026-07-08 23:55 |
- |
|
| #815 |
Fix demo apps hello_world.wait_for_approval + notification_hub.send_alert_and_wait: broken wait_for_event calls (#agentic)
Both call ctx.wait_for_event with the wrong signature: hello_world passes (step_name, timeout=timeou...
|
closed |
medium |
2026-07-07 00:01 |
- |
|
| #814 |
Activity remote context missing 'now' -> non-waiting app with ctx.now fails when routed to activity worker (#agentic)
A non-waiting app with a timeout_policy > 30s routes to the activity worker; if the app calls ctx.no...
|
closed |
medium |
2026-07-06 01:03 |
- |
|
| #813 |
Several apps call ctx.wait_for_event(timeout=...) but SDK AppContext.wait_for_event takes timeout_seconds -> waiting actions crash (#agentic)
AppContext.wait_for_event (enterprise/apps/sdk/context.py:525) signature is (step_name, event_name, ...
|
closed |
medium |
2026-07-06 01:03 |
- |
|
| #812 |
AppToolBridge.is_waiting_tool hardcoded False -> waiting app actions with long timeout mis-routed to activity and FAIL (#agentic)
is_waiting_tool (tool_bridge.py:200) returns False unconditionally (TODO). task.py:80 routes a waiti...
|
closed |
high |
2026-07-06 00:42 |
- |
|
| #811 |
Approval-timeout recovery is flaky under repetition/load (#810 residual, #agentic)
HONEST CORRECTION to my #810 report. #810 FIXED the permanent-hang class (0 infinity, 0 corrupted - ...
|
closed |
medium |
2026-07-05 03:25 |
- |
|
| #810 |
JUMPER-CORE: concurrent claim/suspend/emit leaves runs in corrupted stuck state (#agentic)
Under concurrency (12 parallel approval-timeout runs), runs get stuck: run.state='sleeping' + task.s...
|
closed |
medium |
2026-07-04 01:15 |
- |
|
| #809 |
INVESTIGATE: durable event-wait can hang (approval wake missed under load) (#agentic)
Evidence: in the full harness, a run stayed 'sleeping' 301s after its approval was rejected by the s...
|
closed |
medium |
2026-07-04 00:15 |
- |
|
| #808 |
estimated_cost_usd always 0: agent's Ollama Cloud models missing from PRICING (#agentic)
ROOT CAUSE: engine/tools/llm/usage.py PRICING dict has no 'ollama' provider key. The agent route (co...
|
closed |
medium |
2026-07-03 23:00 |
- |
|
| #807 |
Stage 8 epic: trace UI + docs (#agentic)
Dashboard Agent-trace tab rendering GET /workflows/<id>/trace (agent tree, per-node tokens/cost/dura...
|
closed |
medium |
2026-07-03 22:42 |
- |
|
| #806 |
Stage 7 epic: fork / time-travel (#agentic)
POST /workflows/<id>/fork (+ MCP fork_run): new run of same definition; copy variable+checkpoint row...
|
closed |
medium |
2026-07-03 13:48 |
- |
|
| #805 |
Stage 6 epic: agent progress streaming (SSE) (#agentic)
SQL trigger NOTIFY on highway.agent_event insert (migration; engine untouched -> harness-gated). New...
|
closed |
medium |
2026-07-03 13:24 |
- |
|
| #804 |
Stage 5 epic: HITL v2 - edit-args, timeout, escalation (#agentic)
Replay-safe arg editing: approval decision via durable wait_for_event payload -> merge args={**args,...
|
closed |
medium |
2026-07-03 05:09 |
- |
|
| #803 |
Stage 4 epic: cross-run pgvector memory (#agentic)
Swap postgres:16-alpine -> pgvector/pgvector:pg16 (volume-compatible, demo-gated). Migration CREATE ...
|
closed |
medium |
2026-07-03 04:12 |
- |
|
| #802 |
call_agent agent-to-agent handoff (Stage 3 part 2, #801 task 21)
Implemented call_agent handoff. NEW child template api/dsl_templates/agent_call_child.py (init_goal ...
|
closed |
medium |
2026-07-03 01:37 |
- |
|
| #801 |
Stage 3 epic: agent specs + call_agent handoffs (#agentic)
ALTER highway.agents to carry a full run_goal spec (system_prompt, route, tool_catalog, tool_config,...
|
closed |
medium |
2026-07-02 18:35 |
- |
|
| #799 |
Structured output: bare ollama format constraint yields empty replies from cloud thinking models
First live structured run: glm-5:cloud under ollama native format returned EMPTY content (not valid ...
|
closed |
medium |
2026-07-01 23:24 |
- |
|
| #798 |
Stage 1 epic: structured output through the LLM gateway + Anthropic prompt caching
Agentic plan Stage 1: (a) response_format JSON-schema param through call_llm to all providers (opena...
|
closed |
high |
2026-07-01 23:01 |
- |
|