Commit graph

2 commits

Author SHA1 Message Date
Emi Griffith
64b4eae217 daemon: let the conversational agent own @mentions in server channels
All checks were successful
PR build (required check) / changes (pull_request) Successful in 13s
secrets-guard / encrypted (pull_request) Successful in 9s
shell-lint / shellcheck (pull_request) Successful in 9s
PR build (required check) / build-frontend (pull_request) Has been skipped
PR build (required check) / validate-observability (pull_request) Has been skipped
PR build (required check) / build-backend (pull_request) Successful in 51s
PR build (required check) / gate (pull_request) Successful in 3s
THERMOGRAPH_DISCORD_BOT_MENTIONS=0 makes the gateway bot silent on guild
mentions. Default on, and unset counts as on, so every env file that predates
the knob behaves exactly as before.

The reason is a second responder. The operator's desktop now runs a
conversational agent that replies as this same bot account (discord-voice-bot,
AGENTS.md) and grades through the same API we do, so with both live a single
"@Thermograph Phoenix" gets answered twice — once as a card, once as a reply.
Only one of us can own that surface, and the agent is the one that can also hold
a conversation.

Scoped to guild mentions on purpose. DMs stay here: the agent polls guild
channels only, so staying silent in a DM would just drop the message. And
/grade is a signed HTTP interaction that never touched this path, so it keeps
answering from prod while the desktop is asleep — which is what stops this from
trading an always-on surface for a sometimes-on one.

Claude-Session: https://claude.ai/code/session_015Z1ebLbhUxeZ9ozpNrVTCP
2026-08-02 01:19:01 +00:00
emi
2d3f37c474 daemon: move the Discord gateway and scheduler out of the web process into Go (#21)
All checks were successful
Sync infra to hosts / sync-beta (push) Successful in 8s
Sync infra to hosts / sync-prod (push) Successful in 7s
secrets-guard / encrypted (push) Successful in 8s
shell-lint / shellcheck (push) Successful in 10s
Build + push backend image (Forgejo registry) / build-push (push) Successful in 1m14s
Deploy backend to beta VPS / deploy (push) Successful in 2m4s
The gateway bot and APScheduler were long-lived stateful I/O loops running
inside the async web app under a leader election. They move into a single Go
binary that owns ONLY that I/O -- websocket, RESUME, heartbeat, backoff, timers.

It owns no grading logic. Anything needing data calls back over a new
internal-only surface (/internal/discord/grade, /internal/jobs/*). Grading
depends on polars and the parquet cache; reimplementing it in Go would let the
bot's grades drift from the API's. The grade route returns gateway-ready JSON
and Go relays the bytes verbatim.

The binary ships in the backend image and runs as a second compose service off
the same tag, so the two ends of the /internal/* contract can never skew.
deploy.sh rolls daemon alongside backend -- without that the service would never
be created, since a single-service deploy uses --no-deps. It also probes the
image first and skips the daemon when rolling a tag that predates the binary:
infra tracks main while image tags are env-staged, so a host can legitimately be
asked to roll an older backend image, and creating the service anyway would
leave a container crash-looping on a missing binary.

replicas: 1 with order: stop-first replaces the leader election -- Discord
permits one gateway connection per bot token.

THERMOGRAPH_INTERNAL_TOKEN is optional: both ends derive it from
THERMOGRAPH_AUTH_SECRET via HMAC under a domain-separation label, so this needs
no new vault entry. The derivation is pinned to a shared cross-language test
vector asserted on both sides, so drift fails CI instead of 401ing every call.
Fail closed when neither secret is set.

Improvements over the Python: a close intended for RESUME uses 4000 rather than
1000 (Discord invalidates a session closed 1000, so the old default defeated its
own resume); MESSAGE_CREATE runs on a bounded worker pool; and a malformed HELLO
returns an error rather than a clean reconnect, which would otherwise reset
backoff and hot-loop against the gateway.

365 Python tests pass; Go build/vet/test -race clean; shellcheck 0 findings.
2026-07-23 22:49:54 +00:00