Official issues about message routing, sessions, and platform delivery. Sourced from the official GitHub repo; status derived from official labels; every entry links back to its source with a last-checked date.
send_message_tool's home-channel fallback only copies home.chat_id and drops home.thread_id, so bare `hermes send --to telegram` calls land in the DM root instead of the configured topic when Telegram DM topic mode is active.
The Telegram reconnect handler's updater.stop() call has no timeout, so when the underlying TCP socket is stuck in CLOSE-WAIT the polling task never wakes up, permanently blocking reconnect attempts and silently dropping messages for hours until a manual restart.
Duplicate report of the Telegram cron-job typing indicator persisting for the full subprocess lifetime instead of stopping at message delivery; closed as a duplicate.
A cron job's Telegram delivery keeps the typing indicator active for the entire lifetime of the job's subprocess rather than stopping once the delivery message is sent, which can last 30-60+ minutes for long-running jobs.
Hermes 0.18 persists delivered (post-redaction-restore) conversation text to disk in multiple surfaces (including rich_sent_index.json and the delivery mirror) without any redaction hook or TTL, which is a problem for deployments using PII tokenization plugins that restore PII only at delivery time.
The Telegram gateway adapter hangs indefinitely at connect attempt 1/8 with the 30s asyncio.wait_for timeout never firing, but only under s6-overlay supervision per the reporter's diagnostics.
There is no way to use the assistant directly against a live Excel workbook; the proposal adds an Office.js sidecar with strict tool containment, already implemented per PR #44356.
Local llama.cpp/vLLM-style servers pay a full prefill cost for every new session because the shared prompt prefix is not kept warm; the proposal adds an opt-in gateway watcher that periodically replays a minimal request, and the author notes it is already implemented in PR #57019.
BlueBubbles/iMessage can silently stop delivering inbound messages to Hermes even while process and port signals still look healthy; the issue proposes a real product-level health check instead of relying on weak liveness signals.
## 现象 在 Yuanbao 平台创建 cronjob 时用 `deliver=origin`，投递时好时坏。看日志基本就是两种结果： 1. 正常投递（live adapter 路径命中） 2.
Signal group messages that pass the adapter-level filter are still rejected by _is_user_authorized because Platform.SIGNAL is missing from the platform_group_user_env_map dict used for group authorization.
Setting platforms.weixin.enabled: false in config.yaml does not stop the gateway from connecting to WeChat/iLink when WEIXIN_* environment variables are present. This issue is now closed.
When context compression fires during a /v1/responses conversation chained via previous_response_id, the compressed transcript isn't stored correctly in ResponseStore, so history keeps growing and re-triggers compression on every subsequent request.
After upgrading to v0.18.0, the QQ Bot adapter fails to connect with an unexpected is_reconnect keyword error because gateway/run.py now passes that parameter to all adapters, but the QQ Bot adapter was not updated to accept it.
Turning off gateway.multiplex_profiles does not clean up stale profile sessions in the shared state.db, so the gateway keeps routing messages to a disabled profile.
AsyncSessionDB wraps SessionDB methods as async coroutines, but slash_commands.py calls them without await, causing /resume to crash with a TypeError. This issue is now closed.
A non-interactive gateway run can stay stuck in MCP OAuth discovery for the full browser-callback timeout when token refresh fails, and retries then fail because the callback port is still in use. This issue is now closed.
The context-window usage gauge in the desktop/TUI status bar only refreshes on message.complete, so it stays frozen during long agentic turns even though the underlying data updates after every internal API call.
On macOS, when an external trigger kills the Hermes gateway process, it enters an infinite restart loop of roughly 8-10 second cycles, logging 400+ SIGTERM events in a single session.
With a local Bot API server, Telegram video documents can fail caching with 'method not found', producing empty turns despite the file existing locally.
SlackAdapter checks only whether a session key exists, not whether it would be reset by the configured reset policy, so a Slack thread can lose its expected thread-history reseed after the underlying session actually resets.
This tracking issue consolidates roughly 25 reports of console windows (cmd, conhost, git, gh, powershell) flashing on the Windows desktop GUI when its windowless pythonw.exe backend spawns console-subsystem child processes without the no-window flag, and documents which spawn sites are still leaking based on source and git-history verification.
After updating past a specific commit, the QQBot gateway fails to connect and enters an infinite retry loop, because QQAdapter.connect() receives an unexpected 'is_reconnect' keyword argument.
Since a prior fix (#22773), proactive cron deliveries to a Telegram threaded-DM forum topic land in the General topic instead due to a misclassification in cron/scheduler.py's DM-topic heuristic.
Requests a mechanism to carry conversation context across platforms (Telegram, iMessage, CLI, Discord), since currently only persistent memory carries over and conversation history stays isolated per platform.
Cron deliveries targeting a Telegram private DM topic can fail to stay in that topic because the cron path only carries a generic thread_id, missing the metadata the adapter needs for private DM topics.
This issue proposes adding support for Telegram Bot API 10.1's new Rich Messages (headings, lists, tables, LaTeX, etc.) to the Telegram gateway, replacing the current plain/MarkdownV2 message streaming approach.
Resuming any older chat in the Hermes desktop or TUI app fails with "No LLM provider configured" even though the default provider is valid and new chats work fine, while the identical session resumes correctly via the CLI's --resume flag.
The gateway currently discards Telegram's native forwarded-message metadata (forward_origin, etc.), so the agent cannot distinguish typed text from forwarded messages; the issue proposes a platform-neutral annotation layer to preserve it.
This issue reports that the Codex gpt-5.5 compaction-threshold notice can be shown repeatedly within the same durable gateway session when the AIAgent instance is rebuilt, and proposes showing it only once per session.
Hermes Desktop's Profiles page can list profiles but cannot actually switch which profile a connected remote backend uses; the issue requests a first-class profile switcher in the Desktop UI.
This issue reports that memory operations bypass Hermes's hook system, making tenant isolation impossible without forking core, and proposes upstreaming a fix along with an open-source project called Hermes Swarm Map.
The WhatsApp adapter discards quotedMessageId/context metadata when a user replies to a message, so the agent never receives the quoted message's ID or content, making reply-based corrections unreliable.
On macOS in --tui mode, the gateway repeatedly exits mid-conversation with "stdin EOF (TUI closed the command pipe)" even with only built-in memory enabled (no byterover), which the reporter notes is distinct from a similar SIGPIPE issue (#14036) caused by the byterover memory plugin.
The Teams platform plugin fails to bind port 3978 because it imports the microsoft-teams-apps package, which requires Python 3.12 or newer, while the Hermes installer creates a Python 3.11 venv regardless of a newer system Python being available.
The Matrix/Synapse gateway in the official Docker image breaks after a certain build, with startup logs stalling at 'fixing ownership :1000'; the bot can still send messages but does not respond to channel prompts. Issue is closed as fixed.
The reporter describes a Matrix gateway deployment fronting a custom LLM dispatcher that routes requests between local and cloud models per rule, and reports there is no in-band channel for per-message routing decisions (like a user's /model command) to reach that downstream dispatcher.
This feature request proposes adopting several of Telegram's May 2026 bot platform updates, notably Guest AI Bots (usable via @mention without joining a chat) and Bot-to-Bot communication, to support Hermes multi-agent workflows.
The issue requests adding TrueConf as a supported Hermes messaging gateway, similar to existing Slack/Discord/Telegram support, noting a ready-made python-trueconf-bot library is available.
Slack's mrkdwn does not render GitHub-Flavored Markdown pipe tables, so Hermes responses containing tables displayed as literal pipe text in Slack; the issue is now closed.
User reports that the Hermes dashboard runs as a separate process and does not restart automatically after WSL reboots or gateway crashes, and proposes a systemd service to auto-start it after the gateway.
Feature request for a generic, platform-agnostic API for attaching interactive action buttons (especially Telegram inline keyboards) to any message, instead of hardcoding button support per specific built-in feature.
Feature proposal to add Discord channel history injection and cascade-prevention/mention-gating logic for multi-agent shared-channel setups, based on a working patch the reporter built for their own 3-agent deployment.
This feature request notes that Hermes routes all image analysis through an auxiliary vision model even when the main model is natively vision-capable, and provides a reference implementation (patching 4 files) for native vision bypass along with several related pipeline bugs discovered in the process.
OpenAI Codex OAuth works via CLI, but the Telegram gateway reports 'No Codex credentials stored' despite the same credentials being valid, in a Docker/Coolify deployment. This issue is now closed.
Feature request to let a single Telegram bot route messages from different forum topics/threads to different Hermes profiles (each with its own model, skills, and system prompt), instead of requiring one bot token and gateway process per profile.
Feishu does not render markdown tables sent by Hermes, showing raw text instead, because feishu.py wraps raw markdown in an md tag that Feishu's md tag doesn't support for tables.
This feature proposes a single-daemon architecture, inspired by OpenClaw, letting multiple Hermes agents share one gateway process with isolated per-agent workspaces and memory, routed by session key including per-topic isolation in group chats.
This issue reports that Hermes's Slack adapter uses the legacy mrkdwn format, which doesn't support markdown tables, and proposes switching to Slack's newer Block Kit markdown block type.
Hermes sessions are isolated per platform (CLI, Telegram, iMessage) with no continuity, so work cannot be resumed across platforms; the issue proposes a built-in handoff mechanism to carry state between them.
Hermes currently only supports parallel background tasks via /background, which lacks mid-task interaction and runs sequentially; the issue requests allowing multiple Telegram bots on one account to connect to the same agent with independent sessions.
This issue requests native multi-agent support in Hermes, allowing a single gateway process to serve multiple named agents with isolated sessions, personas, memory, and tool configurations, instead of today's single global agent identity.
This feature request proposes adding a Linear platform adapter to the gateway, allowing users to interact with the agent directly via @mentions in Linear issue comments instead of switching to Slack, Telegram, or Discord.
Hermes cron job results delivered to the origin channel are visible only to the human and not mirrored into the live gateway session's context, so the main agent has no memory of completed scheduled tasks unless a human manually points it to a filesystem handoff file.
This issue documents making Dreaming a first-class, reliable reflection feature across both the CLI and gateway (including Telegram), with tool usage disabled, correct runtime/provider config propagation, and error surfacing instead of silent failure.
This issue reports three mismatches between documented and actual behavior for AGENTS.md and SOUL.md context file loading, including that AGENTS.md is not recursively loaded and cwd is overridden in gateway/messaging mode.
This issue proposes Multi-Role Auto-Routing, letting users define named roles with dedicated sessions and system prompts, auto-routed by a lightweight classifier, to avoid context pollution from mixing unrelated topics in one shared session.
This issue reports that all Hermes gateway sessions across multiple Telegram groups or Discord servers currently share the same personality, prompt, and working directory, and proposes a topic_configs config section for per-chat overrides.
The issue requests adding Rocket.Chat as a supported Hermes messaging gateway channel.
This feature proposes a channel_overrides config section so each Discord/Telegram channel can specify its own model, provider, and system prompt instead of sharing one global model and prompt across all channels.
This issue reports that Hermes Agent's gateway authorization is currently all-or-nothing, and proposes a tiered permission system (Owner/Admin/User/Guest) for messaging platforms.
The issue requested native Matrix protocol support as a Hermes messaging gateway so users could interact via any Matrix client on self-hosted homeservers; the issue is now closed.
Back to curated issues . School solution cards . Troubleshooting