Repository navigation
docs(reactor): fix the consumer example and payload table (SIM-5845) - #173
Conversation
…ent type, acks by signal_id (SIM-5845)
The published loop read .get("signals") (API returns reactor_signals), matched
type == "copytrading" (copytrading signals carry no type; event_type is
"settlement"), and deleted by tx_hash (maker-fill signals use a leg-scoped
signal_id). The field table listed flat taker_* keys that live under whale.
|
Preview deployment for your docs. Learn more about Mintlify Previews.
|
|
The diff is small and contained. Let me verify the simmer-labs FAQ was already fixed (so no companion PR is needed there). The FAQ file at [AI Review] SIM-5845 Verdict: CLEAN Findings: none Summary: All three ticket bugs are correctly fixed — [Codex Review] SIM-5845 Verdict: CLEAN Findings (empty if CLEAN): Summary (one sentence): ready for code-owner review AI review via review_poll.py · Paperclip: SIM-5845 |
This branch was successfully deployed
2 participants
Fixes the public Reactor consumer loop at
/pro/reactor. As published, it never processes a signal:simmermain).get("signals", []){"reactor_signals": [...]}(local_dev_server.py,GET /api/sdk/reactor/pending)type == "copytrading"type;event_typeis"settlement"(reactor_relay._build_signal_payload). Seasonal signals settype(e.g.shock_ladder)tx_hashsignal_id; maker-fill signals use a leg-scoped id (reactor_relay._write_signal)taker_*/market_titlefieldswhale.*The new example matches the shipped
polymarket-copytradingskill (copytrading_trader.py: readsreactor_signals, keeps signals with notype, deletes bysignal_id).Supersedes the docs change Kozy proposed on SIM-5845, whose snippet filtered on
event_type == "copytrading"and would have skipped every signal.Public docs: needs Adrian's approval before merge.