updates for stability and speed - #14
Conversation
🔍 자동 분석 결과이슈 요약업적/배지 시스템 — 사용자 활동에 배지 부여하여成就感 + 수집 욕구 자극 기술 구현 고려사항백엔드 (Phoenix/Ash)
iOS (SwiftUI)
우선순위 제안
관련existing 코드
🤖 HanNyang Auto Resolver |
Task NousResearch#14: FreeBSD Compatibility Audit (partial completion) Changes: - Added FreeBSD to PLATFORM_MAP in agent/skill_utils.py and tools/skills_tool.py - Created comprehensive audit report documenting FreeBSD compatibility issues - Identified voice tools as unavailable on FreeBSD (ctranslate2 has no wheels) - Documented clipboard support gap for FreeBSD - Verified PTY fallback mechanism works when ptyprocess unavailable Files modified: - agent/skill_utils.py: Added "freebsd": "freebsd" to PLATFORM_MAP - tools/skills_tool.py: Added "freebsd": "freebsd" to _PLATFORM_MAP - PLANS/task-coordination.md: Updated task status and added completion notes - PLANS/freebsd-audit.md: Created detailed audit report with findings Next steps: Apply remaining high-priority fixes (clipboard support, setup script) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…on export, pinned sessions, context meter Ported from ibelick/webclaw PRs NousResearch#24, #10, NousResearch#14, NousResearch#13: - Command palette (⌘K): search and switch sessions instantly - Conversation export: download as Markdown, JSON, or Plain Text - Pinned sessions: pin/unpin from context menu, shown at top of sidebar - Context meter: token usage ring in chat header with hover details - Keyboard shortcuts: ⌘K search, ⌘⇧O new session New UI primitives: autocomplete, command, input, preview-card Attachment button/preview components (composer already has built-in support)
…)' (NousResearch#14) from sync/upstream-2026-04-10 into main
The upstream merge (PR NousResearch#14) auto-resolved gateway/run.py's 1598-line diff and silently dropped the entire self-nudge system (~55 lines in gateway/run.py, ~37 in run_agent.py, ~5 in model_tools.py). This broke notify_on_complete for background processes — the mechanism that fires a hidden turn when a background process exits was gone. Cherry-picked b8737bc ("feat(gateway): add one-shot self nudge tool") on top of the merged state, resolving 3 conflicts in gateway/run.py (media_message_callback + self_nudge_callback coexistence) and 1 in tests/test_model_tools.py (kept both clarify + self_nudge tests). The self-nudge system provides: - _arm_self_nudge / _cancel_self_nudge / _fire_self_nudge in gateway - self_nudge_callback on AIAgent for tool-initiated delayed turns - self_nudge tool exposed on gateway platforms (not CLI) - _pending_hidden_turns for injecting hidden messages on next turn - Wired to notify_on_complete in terminal_tool.py via process_registry.pending_watchers
Resolution: Setup Complete — Verification Blocked by Issue #11All setup tasks are complete:
Remaining: Full end-to-end verification of cron firing and Discord output routing. This is blocked by Issue #11 (Anthropic credit must be claimed before LLM calls work for trading skills). Closing this issue — setup is done. Verification will happen automatically once Issue #11 is resolved. No further action needed on this issue. |
…EADME, AGENTS, website Completes the remaining ADDING_A_PLATFORM.md integration points after the MVP vertical slice (ff70b47..3babac1). Fork is now feature- complete per Hermes's own platform checklist; all that's left for an upstream PR is the 1-pt decision on whether NousResearch accepts social-network platforms. Integration points added: - cron/scheduler.py — `"agentchat": Platform.AGENTCHAT` in delivery platform_map so `cronjob(deliver="agentchat:<ch>")` routes - tools/cronjob_tools.py — `deliver` param schema now documents `agentchat:welcome` as a valid target format - hermes_cli/status.py — `AgentChat` row showing `AGENTCHAT_TOKEN` + `AGENTCHAT_HOME_CHANNEL` configuration state in `hermes status` - hermes_cli/gateway.py — setup wizard `_PLATFORMS` entry with interactive prompts for TOKEN / AGENT_ID / ALLOWED_USERS / HOME_CHANNEL / REQUIRE_MENTION / WS_URL (`hermes setup gateway` → pick AgentChat) Docs added / updated: - website/docs/user-guide/messaging/agentchat.md — NEW full setup guide: prerequisites, env vars table, trigger semantics, outbound behavior, MVP limitations, troubleshooting, related npm packages (agentchat-mcp, openclaw-agentchat) - website/docs/user-guide/messaging/index.md — row in platform table + link in Next Steps list - website/docs/reference/environment-variables.md — 8 new AGENTCHAT_* variable rows - README.md — AgentChat added to Messaging Gateway platform list - AGENTS.md — `gateway/platforms/` listing updated channel_directory.py intentionally not modified — AgentChat falls through the default session-based discovery path (`_build_from_sessions`) which is the correct behavior for platforms without native channel-list APIs exposed in the adapter. redact.py intentionally not modified — AgentChat agent_ids are the canonical public handle for messaging (like Matrix `@user:server`), not PII. No regex-level redaction needed. Verified: - 20 unit tests still pass (0.84s) - All 4 modified .py files pass `ast.parse` - Real-world roundtrip previously verified (sha:3babac1 commit msg) Fork state after this commit: 9 of 16 ADDING_A_PLATFORM.md points covered directly, 3 (NousResearch#11 channel_directory, NousResearch#14 redact, NousResearch#16 integration tests) intentionally skipped with justification, 20 unit tests cover the 5 core boundary invariants.
|
Heartbeat-Gestartet um 02:00 - Delegiere an Claude Code Subagent fuer AppViewModel Refactoring (ScanVM, LibraryVM, SyncVM, MetadataVM). Aufwand ~4-6h. |
alt-glitch
added
type/perf
PM had no GitHub tools — claimed PR NousResearch#14 merged to develop but it merged to main. Gave PM github toolset and step 9 to verify baseRefName is 'develop' before declaring implementation complete.
updates for stability and speed
|
Refactor progress update for issue #14. Commit: 6124f6e96 refactor(gateway): move proxy runtime What changed:
Validation:
|
|
Refactor progress update for issue #14. Commit: 6ebe9eaa7 refactor(gateway): move telegram topic rename runtime What changed:
Validation:
|
|
Refactor progress update for issue #14. Commit: 2d6820e6e refactor(gateway): move pending transcription runtime What changed:
Validation:
|
|
Refactor progress update: extracted Telegram topic action behavior from |
|
Refactor progress update: extracted the early inbound control-plane interceptors from |
|
Refactor progress update: extracted the active-session ingress fast path from |
|
Refactor progress update: extracted plugin/bundle/skill slash-command dispatch from |
|
Refactor progress update: extracted cold-path built-in slash command dispatch from |
|
Refactor progress update: extracted user-defined quick command handling from |
|
Command prelude extraction committed as Moved the remaining slash-command prelude out of
Validation:
Line counts after slice:
|
|
Session hygiene extraction committed as Moved pre-agent transcript hygiene/compression out of Validation:
Line counts after slice:
|
|
Session preparation extraction committed as Moved pre-agent session setup out of
Validation:
Line counts after slice:
|
|
Agent input preparation extraction committed as Moved final pre-agent input shaping out of
Validation:
Line counts after slice:
|
|
Post-agent process drain extraction committed as Moved the post- Validation:
Line counts after slice:
|
|
Agent response preparation extraction committed as Moved post-agent response normalization/bookkeeping out of
Kept the source-inspected compression session-id propagation and compression-exhaustion reset blocks in Validation:
Line counts after slice:
|
|
Agent exception handling extraction committed as Moved the
Validation:
Line counts after slice:
|
|
Transcript persistence extraction committed as Moved the post-agent transcript append/update rules out of
Kept the source-inspected compression session-id propagation and compression-exhaustion reset blocks in Validation:
Line counts after slice:
|
|
Agent delivery finalization extraction committed as Moved the final response delivery tail out of
The source-inspected Validation:
Line counts after slice:
|
|
Agent response decoration extraction committed as Moved response decoration out of
Validation:
Line counts after slice:
|
|
Transcript failure classification extraction committed as Moved the context-overflow vs transient-failure transcript classifier out of
Validation:
Line counts after slice:
|
|
Agent turn execution extraction committed as Moved the agent execution envelope out of
Validation:
Line counts after slice:
|
|
Refactor slice completed: progress queue draining Commit: 74c005516 refactor(gateway): move progress queue draining What moved:
Validation:
Line counts:
|
|
Refactor slice completed: streaming/interim setup Commit: e5c72a14e refactor(gateway): move streaming setup What moved:
Validation:
Line counts:
|
|
Refactor slice completed: agent notification callbacks Commit: 4e5acccf9 refactor(gateway): move agent notification callbacks What moved:
Validation:
Line counts:
|
Post-Merge Review SummaryPR: #14 — "updates for stability and speed" What This PR Does
Test ResultsRelevant tests pass (99 total across batch_runner and web_tools test suites):
Review FindingsNo blocking issues found. The PR is already merged and the changes are functional. Minor observations:
P2 Inline Comment (from bot review)The P2 comment about "double-probing terminal readiness" on tools/code_execution_tool.py line 217 is about code that was NOT changed by this PR — the bot scanned the full repo state at the PR's head commit. A fix for that concern was created in commit 8ddafd3 (branch wt/t_a695688a) which makes check_sandbox_requirements() call check_terminal_requirements(). That fix is pending merge to main. Reviewed by Hermes Agent (improver profile) on 2026-07-12 |
7 participants
No description provided.