fix(sessions): export and verify delegate transcripts before cascading session deletion - #71123
Closed
Drexuxux wants to merge 1 commit into
Closed
fix(sessions): export and verify delegate transcripts before cascading session deletion#71123Drexuxux wants to merge 1 commit into
Drexuxux wants to merge 1 commit into
Conversation
alt-glitch
added
type/bug
kshitijk4poor
added a commit
that referenced
this pull request
Jul 25, 2026
Follow-up cleanup for PR #71123: - Extract getattr(args, 'lineage', 'single') == 'logical' to a local (appeared 3x in the export block) - Document that the double _collect_delegate_child_ids traversal in delete_session is an intentional TOCTOU guard inside the write txn
Collaborator
|
Merged via #71187. Your commits cherry-picked with authorship preserved (rebase-merge). The fix exports and verifies every delegate/subagent transcript before cascade-deleting the parent session, with a TOCTOU guard that fails closed if the delegate set changes mid-export. Great catch on a real data-loss bug — thanks for the contribution! |
Th0rgal
added a commit
to Th0rgal/hermes-agent
that referenced
this pull request
Jul 26, 2026
#25) * fix(checkpoints): bind an empty orphan preview to an empty deletion allowlist Follow-up for salvaged PR #69141, addressing the last open review point: cmd_prune() only set orphan_allowlist inside 'if orphans or pre_v2_orphans', so a zero-orphan preview passed the unrestricted None sentinel down to prune_checkpoints(), authorizing deletion of any project that became orphaned between the preview and the rescan — with zero confirmation calls. The allowlist is now bound unconditionally for every non-force run (empty preview => empty allowlist); --force keeps None. Adds the zero-orphan-preview timing regression plus allowlist-identity tests. * fix(agent): cache static system prompt prefixes * fix(prompt-caching): inject cache breakpoints after message normalization The conversation loop normalizes message text right before the API call so the request prefix is byte-identical across turns -- the stated reason is KV cache reuse on local inference servers and better cache hit rates on cloud providers. Cache breakpoints were injected *before* that pass, which defeats it. `_apply_cache_marker` rewrites a plain-string `content` into a `[{"type": "text", ...}]` block. The normalization pass is guarded on `isinstance(content, str)`, so every message that just got marked is silently skipped by it and keeps its raw leading/trailing whitespace. A message is only marked while it sits in the last-3 window, so: turn N in the window -> marked, content "file1\nfile2\n" turn N+1 rolled out -> plain, content "file1\nfile2" The same logical message is sent with different bytes on consecutive turns. The prefix stops matching at that position -- which is inside the span the breakpoints were placed to protect -- so the reusable prefix collapses back toward the system breakpoint on every turn. Tool results carry a trailing newline almost by default (any shell command output), so this is the common case, not an edge case. Move the injection below every message mutation. Besides fixing the whitespace divergence this stops breakpoints from being spent on messages that the orphan sweep or the thinking-only drop is about to remove or merge away -- a marker on a dropped message is a wasted breakpoint out of the four available. Nothing between the old and new call sites reads `cache_control`, and the mutators now see the plain-string shapes they were written against. * fix(caching): reconstruct static system prefix on session restore and post-compression reuse Follow-up to the cherry-picked #68258 base: the cross-session-stable prefix (_cached_system_prompt_static) was only recorded on fresh builds, so two paths silently degraded to the legacy single-breakpoint layout (flagged in review of #68258/#69341/#69704): - Session restore: gateway surfaces build a fresh AIAgent per turn and restore the persisted prompt verbatim from the session DB; the static prefix stayed None from turn 2 onward, flip-flopping the wire layout. - Post-compression cached-prompt reuse: _invalidate_system_prompt() clears the static prefix, and the keep-cached-prompt branch never restored it. Both sites now reconstruct the stable tier and adopt it ONLY when the authoritative prompt string literally startswith() it — stable-tier drift (skills edited, identity changed) falls back to the legacy layout with the stored bytes untouched. Fail-open on any builder error. The restore-path rebuild is gated on _use_prompt_caching so non-Anthropic routes skip it entirely. Refs #68191 Co-authored-by: JonthanaHanh <92574114+JonthanaHanh@users.noreply.github.com> Co-authored-by: joaomarcos <joaomarcosdias444@gmail.com> Co-authored-by: webtecnica <webtecnica@users.noreply.github.com> * fix(config): preserve opaque .env values The .env sanitizer inferred missing newlines from known KEY= substrings inside existing values. Plain secrets containing those bytes could therefore be split into synthetic assignments and rewritten to disk. Treat each physical line as the only assignment boundary and keep bytes after the first equals sign opaque for boundary discovery. Preserve safe formatting, null-byte removal, BOM handling, and normal one-assignment-per-line parsing. Cover direct loading, dotenv loading, sanitization, writers, and migration with behavioral regressions. Fixes #29155 * fix(web): resolve per-profile gateway state for ?profile= in /api/status When ?profile=<name> was passed to /api/status, the handler used _config_profile_scope to set the HERMES_HOME contextvar override, but the gateway liveness check (get_running_pid_cached) and runtime status read (read_runtime_status) both resolve _get_process_hermes_home(), which deliberately ignores contextvar overrides (issue #56986) — it always reads os.environ['HERMES_HOME'] or the platform default. A named profile's gateway identity files (~/.hermes/profiles/<name>/gateway.pid, gateway_state.json) were therefore never found and the endpoint always reported the profile's gateway as stopped. Fix: when ?profile=<name> is requested, resolve the profile directory and pass explicit profile-scoped paths: - get_running_pid_cached(pid_path=profile_dir / 'gateway.pid') - read_runtime_status(path=profile_dir / 'gateway_state.json') - get_runtime_status_running_pid(..., expected_home=profile_dir) This is the same explicit-path pattern _collect_profile_gateway_topology already uses for per-profile gateway state, and it works within the #56986 constraint (no HERMES_HOME env mutation; read-only cross-profile access). Plain /api/status without ?profile= keeps the exact zero-arg calls, so its behavior — including the pid-cache signature and runtime-status fallback — is byte-for-byte unchanged. Fixes #69143 * test(web): pin per-profile gateway state scoping on /api/status Follow-up for the salvaged #70498 fix: replace the original PR's mock-signature churn (28 lambda **kw edits, needed only because it changed the no-profile call shape) with two targeted regression tests: - ?profile=<name> must pass the profile's gateway.pid / gateway_state.json paths and expected_home to the gateway status readers (HOME-anchored per-profile state under ~/.hermes/profiles/<name>/) - ?profile=<unknown> must 404 via _resolve_profile_dir The production change keeps plain /api/status on the exact zero-arg calls, so every pre-existing test passes unmodified. * test: accept the new profile-scoped kwargs in status fakes /api/status?profile= now passes pid_path=/path=/expected_home= to the PID and runtime-status readers; the profile-unification fakes had zero-arg signatures and raised TypeError. Plain /api/status call shapes are unchanged (pinned by the existing zero-arg tests in test_web_server.py). * fix(gateway): prevent reconnect watcher wedge after network-loss fatal error (#70344) Three-part fix for the gateway going silently deaf after a retryable fatal adapter error (e.g. httpx.ConnectError on Telegram): 1. **Detach-on-timeout in _connect_adapter_with_timeout** — Replaced plain asyncio.wait_for with the task-detach pattern used by _await_adapter_cleanup_with_timeout. asyncio.wait_for cancels the overdue task but then waits for it to exit, so a connect() that catches CancelledError can block recovery forever. The detach pattern releases the runner at the deadline via consume_detached_task_result. 2. **Ensure reconnect watcher always runs after escalation** — Added _ensure_reconnect_watcher_running(), called after queueing a retryable fatal error. If the reconnect watcher task has died (exhausted restart budget, terminal exception), it is respawned so queued platforms are never permanently stranded. 3. **Faulthandler at gateway startup** — Enabled faulthandler + SIGUSR2 dump to a rotating file under HERMES_HOME/logs/ for post-mortem diagnosis of future event-loop freezes. Tests added for _ensure_reconnect_watcher_running (alive, dead, not-started, not-running), fatal-error integration (retryable calls ensure, non-retryable does not), and _connect_adapter_with_timeout (timeout raises, success returns). * fix: explicit encoding for faulthandler file open (ruff PLW1514) * fix(gateway): stay alive on mixed retryable + non-retryable startup failures When connected_count == 0 and at least one platform failed with a non-retryable error, the runner exited with GATEWAY_FATAL_CONFIG_EXIT_CODE (78) even if OTHER platforms failed for merely transient reasons. Real-world shape (NS-609, hosted instance): WhatsApp enabled but never paired (non-retryable whatsapp_not_paired) + Telegram TimedOut during polling startup (retryable) => exit 78 => the gateway either goes permanently down (supervisors honoring the exit-78 contract via RestartPreventExitStatus / the s6 finish->125 translation from #51228) or crash-loops (anything else). Either way Telegram never gets its retry and the dashboard drops with every exit, so a single unpaired platform plus one network blip disconnected every channel on the instance. Now exit 78 is reserved for the case where ALL startup failures are non-retryable (true config error, nothing to wait for). With mixed failures the gateway stays alive in degraded state: the reconnect watcher recovers the retryable platforms and the misconfigured ones stay fatal-parked and visible in runtime status. * fix: gate SIGUSR2 faulthandler registration behind POSIX check signal.SIGUSR2 and faulthandler.register() don't exist on Windows; the bare reference raised AttributeError at import time per the windows-footgun checker. faulthandler.enable() still covers fatal-error dumps on all platforms. * fix(gateway): detect and escape silent event-loop freezes - A self-rescheduling 5s call_later floor timer, armed before any adapter connects, guarantees the selector always has a finite timeout, so the existing async defenses (polling heartbeat, timeout guards) regain a chance to run after a zero-pending-timer stall. - A resident daemon-thread liveness watchdog probes the loop via call_soon_threadsafe every 30s; after 3 consecutive 10s-timeout misses (~120s of total unresponsiveness) it dumps all thread tracebacks and exits with the established GATEWAY_SERVICE_RESTART_EXIT_CODE (75) so a supervisor restarts the gateway - async-level recovery cannot run on a frozen loop. - stop() disarms both guards before any teardown await so a busy shutdown is never misjudged as a freeze. HERMES_GATEWAY_LOOP_WATCHDOG=0 disables; _INTERVAL/_TIMEOUT/_STRIKES tune the thresholds. Fixes #69089 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(gateway): close watchdog shutdown race against final-strike exit - Re-check stop_event after a missed probe (before the strike increment) and again on entering the final-strike branch (before the critical log, dump, and hard exit), so a normal stop() landing between the last timeout check and the exit path can no longer be misclassified as a freeze and trigger a supervisor restart. - Deterministic boundary tests pin both re-checks independently (mutation-verified: removing either check turns its own test red); frozen-loop semantics are unchanged. Addresses the shutdown-race review on #69164. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(gateway): recheck stop immediately before watchdog hard exit - A stop() landing while the final diagnostics (critical log, traceback dump) are executing could still reach os._exit(75) after the pre-diagnostic check. Add a third stop_event recheck immediately before the hard exit: diagnostics may complete, but a disarmed watchdog never exits. - Deterministic regressions for both windows (stop triggered from inside logger.critical and from inside faulthandler.dump_traceback); mutation-verified (removing the check turns both red). Frozen-loop semantics unchanged. Addresses the second round of the shutdown-race review on #69164. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * refactor(gateway): gate loop-liveness watchdog via config.yaml, drop HERMES_* env knobs Follow-up to the salvaged #69164 commits: policy forbids introducing new HERMES_* environment variables, so the four watchdog env knobs (HERMES_GATEWAY_LOOP_WATCHDOG / _INTERVAL / _TIMEOUT / _STRIKES) are replaced with a single config.yaml boolean: gateway: loop_watchdog: true # default; false disables both guards - gateway/config.py: new GatewayConfig.loop_watchdog field (default True), parsed from top-level or nested gateway: form, round-trips via to_dict/from_dict. - gateway/run.py: _start_loop_liveness_guards() checks config.loop_watchdog before arming the floor timer + watchdog (getattr-guarded for bare object.__new__ runners). - gateway/shutdown_watchdog.py: start_loop_liveness_watchdog() no longer reads the environment; probe interval/timeout/strikes are module constants (30s/10s/3 — ~90s to restart, matching the systemd watchdog layer's posture). - hermes_cli/config.py: documented gateway.loop_watchdog default so 'hermes config set gateway.loop_watchdog false' validates. - tests: env-knob tests replaced with config-gate + round-trip tests; the final-strike boundary test injects its probe via max_strikes directly instead of patching the removed env helper. * fix: getattr-guard _stop_loop_liveness_guards in GatewayRunner.stop Teardown-path tests build bare runners via object.__new__ without the liveness-guard machinery; the unguarded call raised AttributeError in 8 tests. Same guard pattern as the start path. * fix(desktop): close cross-session leak windows in composer + session refs (#59305) Two React passive-effect timing bugs let a session switch land in the wrong chat: activeSessionIdRef/selectedStoredSessionIdRef (use-session-state-cache) and the composer's attachment-scope swap (use-composer-draft) both mirrored their source props via useEffect, which fires one commit AFTER the new session's view has already painted — a synchronous read/submit in that window observed the outgoing session's ids/attachments. - use-session-state-cache.ts: mirror the session refs synchronously during render instead of a useEffect, guarded to fire only when the prop itself changed (not unconditionally) so an imperative pin from submit.ts / use-session-actions (e.g. a freshly resumed runtime id, intentionally not synced to the source atom) survives an unrelated re-render. - use-composer-draft.ts: the per-thread attachment-scope-swap effect is now a useLayoutEffect, closing the window before paint. - submit.ts / session-context-drift.ts: add a 3rd drift prong comparing the composer's loaded scope (SubmitTextOptions.composerScope) against the submit target, resolved into the same lineage-root domain (resolveComposerSessionKey) the composer itself uses — comparing against the raw tip id would false-positive-abort every submit into any session that has ever auto-compressed. - routes.ts / chat/index.tsx: the primary composer's durable scope key now prefers the route over a possibly-stale store selection (primaryRouteSelectedSessionId). - use-composer-draft.ts: redacted [composer-rehydrate] diagnostic log (counts/kinds/scope only, never raw refs) for future reports in this class. - chat-runtime.ts: normalize attachment id values (url/path) before hashing so a re-attach with a trailing slash or backslash path dedupes correctly. 16 files, 286 tests across the touched/dependent suites (17 files) green, including new regression coverage for each fix. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> * fix(desktop): satisfy eslint import-order rules in use-composer-draft.test.tsx CI's check:lint failed on two perfectionist rule violations introduced by the new test file: type import ordering and missing blank line between the parent-relative and same-directory import groups. No behavior change. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> * fix(agent): prevent shared OpenAI client FD-recycle corruption from stale stream watchdog The streaming stale watchdog was calling _replace_primary_openai_client() from its polling thread, which closes the shared client's connection pool. Worker threads from previous stale-killed attempts may still be unwinding their SSL BIOs, causing TLS application-data to overwrite SQLite file headers via FD reuse. This is the same corruption vector documented in #67142 for Anthropic, where the fix was to never close the shared client from a non-owner thread. Apply the same pattern to the OpenAI-wire path: - Stale stream watchdog: skip shared client replacement - Mid-tool-retry cleanup: skip shared client replacement - Stream retry cleanup: skip shared client replacement The request-local client is already closed via _close_request_client_once. The shared client is replaced lazily by _ensure_primary_openai_client on the next request, which runs on the owning thread. Closes #70773. * fix(agent): retire replaced shared OpenAI clients instead of cross-thread pool close Widen the #70773 fix beyond the three in-request cleanup sites removed in the cherry-picked commit: every remaining path that swaps out the shared OpenAI client could still hard-close its pool from a thread that doesn't own the in-flight sockets (credential rotation/refresh on the turn thread, dead-connection cleanup, gateway cache eviction, transport recovery) — the same FD-recycle corruption vector, just rarer. Add AIAgent._retire_shared_openai_client(): shutdown(SHUT_RDWR) all pooled sockets (FD-safe from any thread, unblocks in-flight readers) but never call client.close() — FD release is deferred to GC, which cannot run until every borrowing thread has unwound its SSL BIO. Refcounting is the ownership handshake; with no borrowers the FDs are released immediately. Wired into: - _replace_primary_openai_client (rotation/refresh/dead-conn cleanup) - try_recover_primary_transport (primary_recovery) - release_clients (gateway cache_evict) agent.close() keeps the hard close: full teardown is a real session boundary where no request may be in flight. Tests: new tests/run_agent/test_70773_shared_client_fd_corruption.py covers the three watchdog/retry sites plus retire semantics; existing close-assertions updated to pin retire-not-close. * test: update credential-refresh tests for retire-not-close contract The three refresh tests asserted the replaced shared client gets close()d — the exact cross-thread close #70773 removes. They now pin the new contract: close() is NOT called from the refresh path; the old client is retired (sockets shutdown, FD release deferred to GC). * fix(doctor): UTF-8/latin-1 fallback when scanning .env Prefer UTF-8 for ~/.hermes/.env provider scans, then latin-1 for cp1252/Notepad files. Add regression test for invalid UTF-8 bytes. * fix: handle non-UTF-8 files in OpenClaw migration script * fix: decode config and state files as UTF-8 on non-UTF-8 locales Several file-I/O call sites still use open() / Path.read_text() / Path.write_text() without an explicit encoding, so they fall back to the platform default. On Windows CN/JP/KR locales (GBK/CP932/CP949) any non-ASCII byte in a config/state/user-content file raises UnicodeDecodeError or UnicodeEncodeError and crashes the caller. to the remaining hot paths: - agent/copilot_acp_client.py: fs/read_text_file and fs/write_text_file (Copilot's read_file / write_file tools, directly reported in #18637 bug 2) - agent/model_metadata.py: context-length YAML cache load + two save sites (context probing is on the call path of every model invocation) - agent/nous_rate_guard.py: cross-session rate-limit JSON state (read + atomic write via os.fdopen) - cron/scheduler.py: user config.yaml read in run_job - gateway/delivery.py: cron output writes for AI-generated content, very likely non-ASCII yaml.dump call sites also gain allow_unicode=True so the emitted YAML preserves non-ASCII chars as-is instead of emitting \u escape sequences. Adds regression tests that monkeypatch builtins.open / Path.read_text / Path.write_text to simulate a GBK locale: each test raises UnicodeDecodeError / UnicodeEncodeError unless the caller explicitly passes encoding='utf-8'. Verified that the tests fail on main and pass with this change, on Linux as well as on Windows. Refs #18637 * fix(cli): add explicit encoding to read_text/write_text calls Path.read_text() and Path.write_text() without explicit encoding default to the system locale encoding. On Windows this is typically cp1252, which causes UnicodeDecodeError for UTF-8 content (JSON configs, user data, service scripts). Add encoding="utf-8" to all read_text() and write_text() calls across 8 CLI files, matching the pattern established in PR #50534 (security_audit_startup.py) and ruff rule PLW1514. Fixed files: - main.py: 4 read_text calls - auth.py: 3 read_text calls - banner.py: 1 read_text + 1 write_text - service_manager.py: 1 read_text + 4 write_text - container_boot.py: 1 read_text + 4 write_text - doctor.py: 3 read_text calls - uninstall.py: 2 read_text calls - gateway.py: 1 write_text call * fix(core,cli,gateway,plugins): add encoding='utf-8' to read_text() calls Path.read_text() without an explicit encoding uses the platform's default encoding. On Windows this is typically cp1252 or mbcs, which causes UnicodeDecodeError or silent data corruption when reading UTF-8 content (JSON files, user text, config with non-ASCII chars). This is the read-side companion to the write_text() encoding fix. Fixed the most critical locations that read JSON data, user content, and config files across 14 files with 31 call sites. Pattern: .read_text() → .read_text(encoding='utf-8') json.loads(path.read_text()) → json.loads(path.read_text(encoding='utf-8')) * fix(install): emit UTF-8 from skills_sync on non-UTF-8 Windows locales On Windows with a non-UTF-8 system locale (e.g. CP936/GBK on zh-CN), Python defaults stdout/stderr to the active codepage. tools/skills_sync.py prints glyphs such as checkmark (U+2713) and up-arrow (U+2191) that GBK cannot encode, raising UnicodeEncodeError mid-run. The installer (scripts/install.ps1) captures this script's stdout and the Rust bootstrap parses it as UTF-8 expecting a JSON result frame. A GBK byte stream (or the traceback it triggers) surfaces as: WARN stdout read error: stream did not contain valid UTF-8 stage=config-templates state=Failed error=install.ps1 -Stage config-templates produced no JSON result frame (exit=Some(0)) i.e. the stage fails even though the script exits 0. install.ps1 already sets [Console]::OutputEncoding = UTF8, but that does not propagate to the python.exe child (Python reads PYTHONIOENCODING / locale, not the console encoding). Fix in two places for defense in depth: - tools/skills_sync.py: reconfigure sys.stdout/stderr to UTF-8 at import so output is valid UTF-8 regardless of caller or active codepage. - scripts/install.ps1: set PYTHONIOENCODING=utf-8 and PYTHONUTF8=1 (scoped to the call, restored afterwards) around the skills_sync.py invocation. * test(install): add UTF-8 regression guard for skills_sync child path Addresses hermes-sweeper review on PR #54866: the installer runs tools/skills_sync.py as a child python.exe whose PYTHONIOENCODING / PYTHONUTF8 the scoped install.ps1 block sets, but there was no regression test for this child-Python UTF-8 path. The existing test_child_process_inherits_utf8_mode covers a different (bootstrap entry-point) flow. Add TestSkillsSyncUtf8Guard: three subprocess tests that import skills_sync (triggering its import-time stdout/stderr reconfigure) and assert the checkmark/up-arrow glyphs the script prints at tools/skills_sync.py:596,675 emit valid UTF-8 and exit 0 even when the child env is left unset or explicitly hostile (gbk). A third test proves the guard is load-bearing by reproducing the crash without it. Also keep the new install.ps1 comment ASCII-only (the checkmark spelled out as U+2713) per the file's PS 5.1 parser-compatibility contract at scripts/install.ps1:79-80; the literal glyph in the comment violated that contract. * fix: add encoding="utf-8" to Path.write_text() calls (P1) Path.write_text() without encoding defaults to system locale encoding. On Windows (cp1252), this silently corrupts non-ASCII content written to JSON files, config files, and cache files. This is the write-side counterpart to the read_text() encoding fix (PR #56115). PLW1514 only covers open() calls — Path methods are unguarded by ruff. 39 instances across 16 files, all passing py_compile. Files changed: - agent/copilot_acp_client.py (1) - tools/web_tools.py (1) - tools/xai_http.py (1) - tools/skills_hub.py (8) - gateway/slash_commands.py (1) - gateway/run.py (5) - gateway/dead_targets.py (1) - gateway/delivery.py (2) - gateway/platforms/qqbot/adapter.py (1) - hermes_cli/gateway.py (1) - hermes_cli/banner.py (1) - hermes_cli/service_manager.py (5) - hermes_cli/container_boot.py (5) - hermes_cli/uninstall.py (1) - hermes_cli/main.py (2) - hermes_cli/profiles.py (3) * fix(hindsight): specify UTF-8 encoding for file I/O on Windows On Windows with CJK locales (e.g. Chinese/GBK), pathlib.Path.read_text() defaults to the system encoding instead of UTF-8, causing UnicodeDecodeError when reading .env or .json config files that contain non-ASCII characters. Explicitly pass encoding='utf-8' to all read_text() and write_text() calls in the hindsight memory provider plugin. * fix(memory): read/write .env as UTF-8 in mem0 and hindsight setup The mem0 and hindsight memory-provider setup routines round-trip the user's ~/.hermes/.env: they read existing lines, update the keys they manage, and rewrite the whole file preserving every other line verbatim. Both used env_path.read_text() / write_text() with no encoding. read_text()/write_text() with no encoding fall back to the system locale (cp1252/GBK on Windows), so on a non-UTF-8 host the preserved lines get mangled or the call crashes on any non-ASCII value, and — because the reader never strips a BOM — a Notepad-edited .env makes the first key fail the in-place match and get duplicated instead of updated. Match the canonical .env readers in hermes_cli/config.py: read with encoding='utf-8-sig' (BOM-tolerant) and write with encoding='utf-8'. mem0/_setup.py already pins utf-8 for mem0.json, so this just aligns the .env path in the same file. Fixes both memory plugins in one class fix. Adds regression tests: a BOM'd .env updates the first key in place (locale-independent, fails without the fix) and non-ASCII existing lines survive the round-trip. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(memory): cover the remaining setup-time .env reads with utf-8-sig Follow-up to review feedback: - mem0 _prompt_api_key read .env with the locale default, so a Notepad BOM hid the first key from the masked current-value lookup; read it with utf-8-sig + errors=replace like the canonical readers in hermes_cli/config.py. - hindsight _load_simple_env used plain utf-8; it also parses the Hermes .env during post_setup, where a BOM stuck to the first key. Switch to utf-8-sig + errors=replace. - Add hindsight regressions: BOM key matching in _load_simple_env and in the cloud post_setup writer, plus non-ASCII round-trip preservation, and a mem0 regression for the BOM'd masked-key lookup. The BOM tests fail without the fix on any platform. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(profile): read .env as utf-8-sig in the distribution-install preview `_render_distribution_plan` reads the target profile's `.env` to decide whether a required env var is already set (so it doesn't nag the user), using `Path.read_text()` with no encoding. Two bugs: 1. `Path.read_text()` defaults to the system locale (cp1251/GBK on Windows), which raises `UnicodeDecodeError` on any non-ASCII byte. The surrounding `except OSError` does NOT catch that — `UnicodeDecodeError` is a `ValueError` — so a mis-encoded `.env` aborts the entire install preview. 2. Even on a UTF-8 locale, a Notepad-added BOM prefixes the first key (`KEY`), so the very first required env var is mis-reported as "needs setting" when it is actually present. `.env` is written as UTF-8 everywhere in the codebase. Read it as `utf-8-sig` (tolerates the BOM) and also catch `UnicodeDecodeError` so a genuinely un-decodable file skips the pre-check instead of crashing. Regression tests: a BOM-prefixed `.env` whose first key must still read as "set", and an invalid-UTF-8 `.env` that must not abort the preview. * fix: add UTF-8 encoding to read_text/write_text in tools/ and agent/ Path.read_text() and Path.write_text() without encoding= default to the system locale (cp1252 on Windows), which corrupts non-ASCII JSON content. Coverage-gap fix for files not addressed by prior encoding PRs: - tools/skills_hub.py: 6 read_text + 8 write_text (cache, index, lock files) - tools/skills_sync.py: 1 read_text (lock file) - tools/xai_http.py: 1 read_text + 1 write_text (auth store, marker) - agent/shell_hooks.py: 1 read_text (allowlist) - gateway/status.py: 1 read_text (PID file) - hermes_cli/banner.py: 1 read_text + 1 write_text (update cache) All sites read/write JSON or short text. No behavioral change on Linux (already UTF-8); fixes silent data corruption on Windows. * fix(skills): tolerate non-UTF-8 bytes in hub lock.json _read_hub_installed_names() reads ~/.hermes/skills/.hub/lock.json with a strict utf-8 decode. Hub skill descriptions can carry Windows-1252 typographic bytes (em-dash 0x97, smart quotes, bullets) as single high bytes; read_text(encoding="utf-8") then raises UnicodeDecodeError, which is a ValueError sibling not caught by the function's except (OSError, json.JSONDecodeError). It escapes and 500s the whole /api/skills endpoint, blanking the desktop Skills panel. Decode with errors="replace" so the offending byte degrades to U+FFFD and the structurally valid JSON — and every other skill — stays readable. Fixes #68053 * fix(cli): read .worktreeinclude and .gitignore as UTF-8 in worktree setup _setup_worktree read both files with the locale default encoding. On a cp1251/GBK Windows machine a UTF-8 include list either decodes to mojibake paths (non-ASCII entries silently not copied) or raises UnicodeDecodeError, which the enclosing handler logs at DEBUG and swallows — no include is copied at all, so the worktree starts without .env/keys and the agent breaks invisibly. A Notepad BOM likewise glues to the first include entry on every platform, and to the first .gitignore line, defeating the '.worktrees/' membership check and appending a duplicate entry on each run. Read both files with utf-8-sig + errors=replace, matching the canonical .env readers in hermes_cli/config.py (utf-8-sig because Notepad adds a BOM) and the UTF-8 append this same block already performs on .gitignore. Regression tests exercise the real cli._setup_worktree: the two BOM tests fail without the fix on any platform, the non-ASCII include test additionally reproduces the Windows locale failure. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(skills): read OOXML parts as bytes and form JSON as UTF-8 in office skill scripts The bundled office skills (#68595) read user documents and agent-authored payloads with the locale-default codec: - docx/powerpoint validators/base.py opened OOXML part XML in text mode before handing it to lxml. On Windows (cp1251/GBK) the bytes decode to mojibake that lxml then parses, so validation runs against silently corrupted document text; on locales where the UTF-8 bytes don't decode the validator crashes with UnicodeDecodeError instead of validating. Opening as bytes lets lxml honor the encoding declared in the XML prolog. - The pdf form scripts (fill_fillable_fields, fill_pdf_form_with_annotations, create_validation_image, check_bounding_boxes) read the fields JSON the agent authors — UTF-8 by construction — with the locale codec, so non-ASCII form values (any Cyrillic/CJK/accented input) either crash or get written into the user's PDF as mojibake. The json.dump writers use ensure_ascii=True and were already safe; only the readers needed pinning. Adds a contract test asserting every document/payload reader is locale-independent, plus a live regression test that runs check_bounding_boxes.py on a non-ASCII fields.json under a forced non-UTF-8 locale — it fails without the fix on both POSIX (C locale) and Windows (cp1251 chokes on the 0x98 byte of U+2018). * fix(windows): sweep remaining bare read_text/write_text sites + linter rule AST-driven pass over every Path.read_text()/write_text() without an explicit encoding= across non-test code: 71 sites in 34 files (skills_hub, hermes_cli/main+profiles+service_manager+container_boot, mem0/hindsight/honcho plugins, achievements dashboard, release/CI scripts, productivity+comfyui skill helpers, agent/*). Verified zero positional-encoding collisions before insertion; per-file compile() check after. Adds a check-windows-footguns rule flagging bare single-line read_text/write_text (multi-line forms stay covered by the AST guard test from #38985). Together with the salvaged contributor commits this retires the ~169-site bare file-I/O class (#37423's long tail). * fix: restore utf-8-sig BOM tolerance at .env readers the sweep normalized The cherry-pick auto-resolution + AST sweep applied plain utf-8 at three .env reader sites where the salvaged PRs (#62617, #62123) deliberately use utf-8-sig — a Notepad BOM must not hide/duplicate the first key. Restore the contract (tests pin it). * chore: contributor email mappings for the file-I/O salvage * refactor(desktop): add shared Field form-dialog primitive Dialog forms each hand-rolled their own label+control+hint stack (or borrowed the settings-surface ListRow), so gaps and hint styling drifted between the profile, cron, and webhook dialogs. Add a single Field / FieldHint primitive for label-over-control dialog fields and adopt it in the create/rename profile dialogs as the first consumers. * fix(desktop): unify overlay-pane padding and add primary PanelAction Overlay panes each set their own top padding, so the Settings sidebar and Panel headers sat at different heights than System/Agents and the close X (the #67759 regression). Hoist the shared beside-the-X clearance into OVERLAY_TOP_CLEARANCE, keep the taller pad only on OverlayMain (which sits under the X), tighten OverlayMain's gutters, and drop the one-off Settings override. Also give PanelAction a `primary` variant so a detail header can promote its main action to a filled button. * refactor(desktop): fold cron Blueprints into the New Job dialog Blueprints lived behind a separate Jobs/Blueprints tab with its own card gallery — a bespoke surface no other overlay uses. Remove the tab and make blueprints a "Start from" dropdown at the top of the New Job dialog (default "Custom" = the manual editor); picking one swaps the form for that blueprint's typed slots. Also promote the detail-view "Trigger now" button to a primary action and adopt the shared Field primitive. * refactor(desktop): webhooks create form uses shared Field; drop status pill The create dialog used the settings-surface ListRow/ToggleRow inside a modal, which read differently from every other form dialog, and the detail header carried an enabled/disabled pill that rendered as a stray dash. Switch the form to the shared Field primitive (+ Switch) and remove the pill. * fmt(js): `npm run fix` on merge (#71099) Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> * fix(dashboard): add lightweight /api/health liveness endpoint /api/status is the only public liveness route, and its handler loads the gateway config, probes gateway health, and counts sessions before it can answer. That work is wrong for a readiness probe: a caller that only needs to know the process is up pays for a cold plugin import tree. Add /api/health, which returns process liveness, version, and the auth-gate shape and touches nothing else. * fix(desktop): probe /api/health for boot readiness, and survive a stalled loop Desktop boot polls /api/status, so readiness waits on gateway config and a cold plugin import tree. On Windows that regularly outlives the probe and Desktop kills a backend that is already listening, respawns it, and re-pays the same import cost — the reported crash loop. Probe /api/health instead, falling back to /api/status only for the missing-route shapes the fetch helpers emit (404, or HTML from the SPA), so an older remote backend still connects. Timeouts and server errors keep polling health rather than dropping to the heavyweight route. A cheap route is not enough on its own. Warming the gateway import holds the GIL, so the event loop can stall for tens of seconds and starve /api/health too. At the default 15s socket timeout only three attempts fit in the 45s budget; give each probe 5s so the loop keeps retrying across the stall. Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com> Co-authored-by: DESXIE <78300229+DESXIE@users.noreply.github.com> Co-authored-by: frohsinnllc <231045016+frohsinnllc@users.noreply.github.com> * fix(state): decode display_metadata at every message read path get_messages(), get_messages_around() and get_anchored_view() returned the raw display_metadata column instead of the dict every caller expects. The desktop paints a resumed transcript from the REST prefetch, which reads through get_messages(), so any session holding an async_delegation_complete event failed resume with "Cannot use 'in' operator to search for 'task_count'" — on every such session, not just corrupted ones. Route all four read paths through one shared codec that also unwraps rows carrying a second JSON layer, so sessions already broken on disk recover on read rather than needing a migration. Co-authored-by: Studio729 <Studio729@users.noreply.github.com> Co-authored-by: aml1973 <aml1973@users.noreply.github.com> Co-authored-by: xxxigm <xxxigm@users.noreply.github.com> * fix(state): stop double-encoding display_metadata on write export_session() reads through get_messages(), so before the read fix an already-serialized string went straight back into _insert_message_rows() and got re-dumped — an export/import round trip permanently corrupted the row. Guard the three write paths the same way tool_calls already is: parse a string argument before storing it, and drop metadata that isn't an object rather than persisting something no reader can use. Co-authored-by: xxxigm <xxxigm@users.noreply.github.com> Co-authored-by: aml1973 <aml1973@users.noreply.github.com> * fix(desktop): tolerate unparsed display_metadata from an older backend The desktop and the Hermes backend it talks to version independently — a remote VM running an older build still serves display_metadata as JSON text. Indexing into that string with `in` threw and failed the whole resume, so narrow the type to admit a string and parse it before reading task_count. Falling back to the generic label keeps a delegation event renderable even when the metadata is unusable. Co-authored-by: xxxigm <xxxigm@users.noreply.github.com> Co-authored-by: Studio729 <Studio729@users.noreply.github.com> * fix(checkpoints): don't prune a project whose volume is merely unmounted Orphan pruning decides a project is gone from a single probe: if delete_orphans and (not workdir or not Path(workdir).exists()): reason = "orphan" then deletes its ref, index, and metadata — the project's entire checkpoint history. `Path.exists()` is False for a deleted directory, but it is equally False for one whose storage is not attached right now: an unplugged external drive, a share behind a downed VPN, a bind-mount absent from this container, an offline Windows mapped drive. The project is fine; only our view of it is. This is not an opt-in maintenance command. `maybe_auto_prune_checkpoints` runs unattended at startup from both `cli.py` and `gateway/run.py`, with `delete_orphans=True` by default. So starting Hermes once while the drive is unplugged silently destroys the restore points for every project on it — the one thing checkpoints exist to provide, and there is nothing to restore from afterwards. Reproduced against the real store: a project registered under an unmounted path and one on local disk, then a startup prune — prune: {'scanned': 2, 'deleted_orphan': 1} unreachable project index still on disk: False The legacy pre-v2 branch has the same flaw plus a second one: a `HERMES_WORKDIR` marker that exists but cannot be read leaves `workdir = None`, which the same condition treats as an orphan. Failing to read a file is not evidence that a project was deleted. Require corroboration before deleting: the workdir's parent must be present, so its absence is something we actually observed. A missing parent means the volume is not there and we know nothing, so the entry is left alone — and an unreadable marker never deletes at all. Genuinely abandoned projects are still reclaimed, both by the unchanged orphan path (parent present, project gone) and by the retention/stale rule, which runs off `last_touch` rather than a filesystem probe. tests/tools/test_checkpoint_manager.py: a project whose whole mount disappears keeps its history; controls prove a genuinely deleted project is still pruned and a live project is untouched. The data-loss test fails on main; both controls pass there. 81 passed across the checkpoint suites (2 failures in test_checkpoint_manager.py are pre-existing and fail identically on clean main). * fix(checkpoints): an empty surviving mount point is not evidence of deletion Addresses @egilewski's review: the parent-directory check still deleted checkpoint history for the most common unmount layout. Detaching storage removes the parent outright in some layouts (`/Volumes/Ext/proj` on macOS, `/media/<user>/<label>/proj`), which the first commit handles. But in the classic static layout — `/mnt/volume/proj`, an fstab entry, a container bind-mount — unmounting removes the contents and leaves the mount point behind as an empty directory. `parent.is_dir()` is then true, the project is absent, and the startup sweep deletes its ref, index and metadata: exactly the case this PR set out to protect. Reproduced against the real predicate before this commit: mount root vanished (macOS) -> False ok empty surviving mount point -> True <-- history deleted really deleted (siblings) -> True ok An empty parent carries no information: it looks identical whether the volume was detached or the project was deleted. So require the parent to actually say something — it holds some other entry (we observed a populated directory that does not contain the project), or it is itself a live mount point (the volume is attached right now and demonstrably does not hold the project). The cost is that a project deleted out of an otherwise-empty parent is no longer reclaimed by the orphan rule. It is not leaked: the retention rule reads `last_touch` rather than probing the filesystem and still collects it, so reclamation is deferred, not lost. That is the right direction for a predicate whose false positive destroys a user's restore points unattended. `_dir_has_any_entry` stops at the first entry via `os.scandir` instead of materializing a listing, since a project root can hold a large tree. tests/tools/test_checkpoint_manager.py: `test_surviving_empty_mountpoint_ keeps_its_checkpoints` pins the reviewed case, and `test_empty_parent_project_ is_still_reclaimed_by_retention` pins the deferral above so the safety valve cannot silently regress into a leak. Both fail on the previous commit. The real-orphan control now seeds a sibling so it exercises a populated parent rather than the ambiguous empty one. 80 passed in the checkpoint suite; the 2 remaining failures (`TestGitEnvIsolation`, `TestClearFunctions`) fail identically on clean main. * fix(checkpoints): require positive volume-attachment evidence before orphan classification Follow-up to the cherry-picked #69063: egilewski's review found that the _dir_has_any_entry(parent) guard treats ANY entry in the mount point's parent as proof the volume is attached — but unmounting exposes the UNDERLAY directory's own files (e.g. a .keep placeholder), so a populated underlying mount-point dir still classified the project as an orphan and deleted its ref/index/metadata. Reproduced on both main and the PR head. Attachment evidence is now positive instead of circumstantial: * _volume_evidence() records the parent directory's (st_dev, st_ino) identity in the project's metadata while the workdir is observably live (at _register_project/_touch_project time). A mount point resolves to the mounted filesystem's root while attached and to the underlay directory after detach — same path, different directory, different identity. * _workdir_is_observably_gone() now requires the parent visible at prune time to match that recorded identity before the populated-parent check can classify an orphan. A mismatch means a different directory (the underlay) is showing through — a detached volume, not an observed deletion. * Metadata without a recorded identity (written by older versions) is never orphan-classified — unsure never deletes; the retention/stale rule still reclaims genuinely abandoned projects off last_touch. * The frozen pre-v2 layout has no metadata channel for the identity, so it keeps the structural checks only (require_parent_identity=False). * A failed evidence probe on re-registration preserves the previously recorded identity — stale evidence can only make pruning MORE conservative. Windows: st_dev/st_ino of 0 (filesystems without file IDs, some network shares) is treated as "no evidence recorded", which falls into the conservative never-orphan path. os.path.ismount and Path.stat are cross-platform; no POSIX-only calls added. tests/tools/test_checkpoint_manager.py: adds egilewski's exact regression (checkpoint history for mnt/volume/project, detach exposes mnt/volume/.keep, prune with orphan deletion enabled → NOT deleted; fails on the bare cherry-pick, passes with this fix), plus no-recorded-identity conservatism and probe-failure identity preservation. His absent-parent/empty-parent/retention/genuine-deletion/ live-project controls all still pass. Reported-by: egilewski (review on #69063) * fix(telegram): require initial polling readiness Use wall deadlines for deleteWebhook and start_polling, then fail cold startup unless getUpdates proves progress. This lets the gateway discard partial PTB state and retry with a fresh adapter.\n\nRefs #67498 * fix(gateway): allow Telegram readiness budget Give Telegram a 180s default outer connect budget so cold polling can prove getUpdates readiness. Preserve the 30s default for other platforms and all explicit config/env overrides.\n\nRefs #67498 * fix(telegram): bind strict cold-start readiness to its own polling generation Follow-up hardening for the salvaged #69240 readiness gate (#67498): - _start_polling_once now returns its (generation, progress_event) pair so the strict cold-start gate binds to exactly the generation it started, instead of re-reading self._polling_progress_event which a concurrent recovery task may have replaced with a newer generation's event (the G1/G2 race flagged in the #69240 review). - Strict cold start no longer schedules background polling recovery: a polling error during the readiness wait is captured by a strict callback and fails the connect attempt immediately with a loud OSError, so GatewayRunner disposes the partial adapter and retries with a fresh one — no more waiting out the full readiness deadline on a generation that already errored, and no G2-on-partial-app healing. - After readiness is proven the strict callback delegates every later polling error to the real background-recovery callback, preserving the existing degraded/reconnect semantics for the polling lifetime. - The readiness-timeout error message now states the deadline and that the gateway will retry with a fresh adapter (loud failure, not a silent wait). - Regression tests: current-generation progress connects; a polling error during strict cold start fails fast without scheduling background recovery (the #67498 idle-threads shape); stale-generation progress is rejected. Progresses #67498 * test: record getUpdates progress in mocked cold-connect polling flows The strict cold-start readiness gate (#67498) means adapter.connect() no longer returns True until the mocked start_polling records a successful getUpdates round trip for its generation. Update the conflict-suite Application mocks accordingly: - fake_start_polling side effects call adapter._record_polling_progress(adapter._polling_generation) on the initial connect (retry generations intentionally do NOT auto-progress where a test asserts the conflict count survives an unproven retry). - _build_polling_app takes the adapter so its start_polling mock can record progress. Without this, the cold connects in these tests wait out the full 60s readiness deadline and fail — which is exactly the fail-closed behavior the gate is supposed to provide when polling shows no progress. * fix(config): add a collision-safe env var name for custom endpoint keys Both the Desktop panel and the CLI setup flow need somewhere in .env to put a custom endpoint's API key. Deriving the name from the endpoint's hostname collapses two servers on one machine onto a single slot, and every IP-based local endpoint slugs to a digit-leading name that save_env_value rejects outright. Key off the endpoint's own identity and keep a fixed prefix. Co-authored-by: asorry75 <33794789+asorry75@users.noreply.github.com> Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com> * fix(windows): verify rebuilt Hermes.exe integrity before shipping it as an update (#69179) The desktop self-update chain (Desktop -> hermes-setup --update -> hermes update -> hermes desktop --build-only -> relaunch) rebuilds Hermes.exe on the user's machine and declared success on bare file EXISTENCE. A truncated PE (corrupt cached Electron zip / interrupted extraction or rcedit rewrite / full disk) or a wrong-architecture unpacked tree therefore shipped as the 'updated' app, which Windows refuses to load with 'This app can't run on your computer' (此应用无法在你的电脑上运行) — and the previous working build had already been wiped by before-pack.mjs, leaving nothing to fall back to. Fix, in three parts: - hermes_cli/main.py: post-build integrity gate on Windows (_ensure_desktop_exe_launchable). Parses the PE header of the freshly built Hermes.exe — MZ/PE magic, section-table completeness vs file size (catches truncation), and COFF machine vs the host arch (catches arm64/x64 mixups). On failure it purges the (likely corrupt) cached Electron zip, invalidates the content-hash build stamp so the updater's retry-once genuinely re-downloads and rebuilds, restores the previous build from the .bak tree when one exists (keeping the corrupt tree as .corrupt for diagnostics), tells the user the update was aborted and their old version kept, and exits nonzero. _desktop_packaged_executable also now prefers a host-loadable PE over pure newest-mtime when multiple win-*-unpacked trees coexist. - apps/desktop/scripts/before-pack.mjs: on win32, the previous unpacked tree is preserved as <appOutDir>.bak (only when it holds the product exe — partial/corrupt trees still get the plain wipe) instead of being destroyed, providing the rollback material for the gate above. Non-Windows behavior is unchanged. - Behavior-contract tests: tests/hermes_cli/test_desktop_exe_integrity.py (23 tests — synthetic PE fixtures for truncation/non-PE/arch-mismatch, rollback semantics, and the build-only exit contract) and 6 new vitest cases in before-pack.test.mjs for the .bak preservation rules. Progresses #69179 * fix(desktop): persist the whole discovered model list when saving an endpoint Test enumerates a custom provider's catalogue and the panel holds the result in discoveredModels, but the save payload never carried it, so only the one model the user hand-typed reached providers.<id>.models. Every downstream picker reads that map straight from config.yaml with no live probe, which is why a proxy serving 18 models offered exactly one. Send the discovered list and merge it onto the entry, so models already known keep their context lengths. Fixes #69988 Co-authored-by: asorry75 <33794789+asorry75@users.noreply.github.com> * fix(web_server): keep Desktop custom endpoint API keys out of config.yaml The Custom Endpoints panel wrote the raw key to providers.<id>.api_key, so the credential sat in plaintext in a file users routinely share and commit. The input is masked, so nothing warned them. Write the key to .env and reference it via key_env, the same indirection built-in providers use and that runtime_provider already resolves. The read side has to move with it: reporting has_api_key from api_key alone would show "no API key" for every migrated endpoint, and activate copying only api_key would drop the credential entirely. Delete now clears the .env slot too, and an entry still carrying a pre-fix plaintext key is migrated on its next save so existing users get cleaned up without re-entering anything — unless the key is a hand-written ${VAR} template, which is already safe and must not be duplicated into a second env var. Fixes #69449 Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com> Co-authored-by: asorry75 <33794789+asorry75@users.noreply.github.com> * fix(cli): store custom endpoint API key in .env instead of config.yaml hermes model's custom-endpoint flow is the other write path that produced a plaintext key, on both the model block and the custom_providers entry. Route it through the same .env indirection as the Desktop panel, and swap an existing entry's inline key for the reference when the URL is re-saved. Co-authored-by: liuhao1024 <sunsky.lau@gmail.com> * test: cover custom endpoint key storage and model-list persistence Bug-class coverage for both fixes: the full catalogue survives Save, context lengths are preserved, the key never lands in config.yaml on either write path, blank clears it, a pre-fix plaintext key migrates while a ${VAR} template is left alone, two endpoints on one host keep separate credentials, and an IP-derived name is still a valid POSIX env var. The two delete tests asserted on the plaintext mirror; they now assert the same invariants against the credential reference. * fix(desktop): persist @image: refs instead of the vision-enrichment text The desktop gateway passed the vision-enriched, model-only message text (carrying an `image_url:<path>` hint) straight into run_conversation as the persisted user turn. The renderer only parses `@image:<path>`, so it could not rebuild the attachment from history: after a restart the image was gone and only the caption survived, and on a live session switch the warm cache disagreed with the authoritative text and the frontend "rescued" the image by appending it after the caption. run_conversation already supports persist_user_message for exactly this "what the model sees" vs "what gets stored" split; it was simply never wired up for the attachment path. * fix(desktop): keep cached attachment refs on session resume Persisted history carries no attachment metadata for non-image refs, so resume reconciliation dropped `@file:` chips off a user turn whose text matched. Carry the warm cache's refs forward when the resumed message has none of its own, never replacing refs that are already present. (cherry picked from commit eac5b0a8ac39eab242a5d571531e386ec70e2435) * fix(desktop): quote persisted @image: paths so spaced paths render The unquoted alternative in the directive pattern is `\S+`, so a ref built by string interpolation truncates at the first space and strands the tail as loose text next to a broken thumbnail. Composer images live in the app's userData dir, which on macOS is `~/Library/Application Support/<App>/` — so every pasted or dropped image hit this. Adds format_reference_value next to REFERENCE_PATTERN, mirroring formatRefValue in the desktop's directive-text.tsx, and covers the round-trip through the parser. * fix(desktop): persist the image ref for natively-vision-capable models too A turn routed to a model that takes pixels directly sends `content` as a parts list, and the session store deliberately ignores a plain-string persist override for a list payload — a text override must not erase a turn's image summary. So the override was dropped for every user on a vision-capable main model, and the durable row kept only the caption plus a literal `[Image attached at: ...]` / `[screenshot]`, which the renderer cannot turn back into an image. Only vision-preprocessed (text-mode) turns were actually fixed. Mirror the shape instead: swap the text part for the `@image:` ref form and keep the image parts, so the model still has the pixels for the rest of the session, and drop the `[screenshot]` stand-in on the way into the bubble when a ref was lifted from the same message. * refactor(desktop): memoize the directive image-segment filter Matches the two derived values above it and fixes the indentation. * fix(desktop): lead persisted image turns with the caption Session previews are the first 60 characters of the first user message, so persisting the @image: directives ahead of the caption labelled the session with a truncated file path in the sidebar, session switcher, and command palette. Clients lift the refs out of the body line by line, so moving them after the caption changes nothing about how the turn renders. * test(desktop): cover attached-image resume end to end The unit tests cover each layer in isolation, but nothing exercised the whole chain the bug lived in: the real gateway persisting an attachment, SessionDB holding it after the process exits, and the renderer rebuilding a thumbnail from the stored turn. Seeds a session through the real gateway with an image attached, then launches desktop against it — so the first render is already the relaunch case. Pins native image routing (the majority path, and the one where a text-only persist override is dropped) and stages the file behind directory and file names with spaces, mirroring the macOS composer's Application Support path. * fix(models): resolve custom provider model ids Map picker-prefixed custom provider selections back to their configured model IDs before validation, persistence, and API requests. Fixes #68347 * chore(contributors): map jevin@jevin.org to ijevin Attribution check needs a mapping for the cherry-picked commit's author so release notes credit them correctly. * fix(relay): normalize forwarded Discord interactions to leading-slash commands (#71048) A real APPLICATION_COMMAND interaction forwarded over the relay arrived slash-less: _discord_interaction_to_event set text = data['name'] ("new", not "/new"), MessageType.TEXT, and dropped options entirely — so a registered /new dispatched as plain chat instead of a command (MessageEvent.is_command() is text.startswith("/")). Port the connector's Slack slash-command precedent (normalizeSlackCommand builds `${command} ${args}`.trim() with a leading slash and explicit command type): for type-2 interactions build "/" + name, append rendered options space-separated (scalar options contribute their value, matching the native adapter's f"/model {name}" shape; SUB_COMMAND/ SUB_COMMAND_GROUP contribute their name then recurse into nested options), and set MessageType.COMMAND. Type-3 (custom_id) and other interaction types are unchanged. This implements the interaction->command sub-design previously flagged as deferred in the _on_passthrough docstring. Companion connector fix in gateway-gateway: fix(relay): strip own-mention prefix so addressed slash commands dispatch. * feat(desktop): add session link title resolver Resolve @session:<profile>/<id> reference values to the session's title: the in-memory sidebar list answers most lookups, and an unknown id falls back to GET /api/sessions/{id}. Cache, in-flight dedupe, and subscriber fan-out mirror the external-link title resolver. An untitled row resolves to empty rather than "Untitled session" so the caller's short-id fallback stays the chip label. * feat(desktop): show resolved titles on @session chips Route session refs in the transcript through the title resolver so a dropped session reads as its title instead of a truncated id, and use Tabler's funnel for the session chip icon. * feat(desktop): render agent-written @session links as chips Assistant text goes through the markdown renderer, not DirectiveContent, so a session reference an agent wrote came out as literal text. Rewrite bare refs into `#session/<value>` links during markdown preprocessing and dispatch that href to the shared chip in MarkdownLink, alongside the existing media and preview hrefs. Preprocessing already skips code fences and inline code, so a ref being discussed in code stays literal. The pure parsing/href helpers move to session-refs.ts to keep the resolver's React and API imports out of the per-flush preprocess path. * fix(sessions): export delegate cascade before deletion * refactor: extract lineage_is_logical local + document TOCTOU re-query Follow-up cleanup for PR #71123: - Extract getattr(args, 'lineage', 'single') == 'logical' to a local (appeared 3x in the export block) - Document that the double _collect_delegate_child_ids traversal in delete_session is an intentional TOCTOU guard inside the write txn * feat(session-search): give the agent a link to hand back Asked to link to a session, the agent had no way to know the @session reference syntax exists — every mention in the tool schema described consuming a link the user dropped, never writing one — so it answered with the title and timestamp as prose and the desktop had nothing to render. Every result now carries a ready-to-copy `link`, and the schema says to write it inline instead of restating the title around it. The profile segment is omitted when the active profile can't be named confidently; a bare id still resolves. Also skip linkifying a ref a model already wrapped in a markdown link, which would otherwise rewrite into a nested link. * fix(tui_gateway): retain failed turns as replayable inflight snapshots A turn that ended in error cleared inflight_turn and emitted its terminal frame in the same breath. If the client was disconnected during that window (the exact case for a failure like a network drop), the frame went to the detached drop-transport and the in-memory state was already gone — the desktop reconnected to a session with no trace of the failure. Failed turns now retain a compact error snapshot (user prompt, partial assistant text, error, recoverable) that session.resume's inflight payload carries to a reconnecting client. Covers all three loss sites: the returned-error result path, the turn exception path (which now closes with the same status:"error" message.complete frame shape instead of a bare error event), and agent-init failure. The snapshot lives until the next turn starts or the session closes; _run_prompt_submit replaces a retained error leftover instead of appending onto it. Co-authored-by: Reza Sayar <rsayar@uvic.ca> * feat(desktop): crash-survivable in-flight turn journal The renderer's session-state cache is memory-only and the backend's inflight snapshot dies with the backend process, so nothing survived a full app or machine death mid-turn: reopening the session showed the transcript up to the last committed turn and silently dropped everything the crashed turn had streamed. While a turn runs, the visible tail (user prompt + streamed assistant rows, tool calls included) is now journaled to localStorage — throttled off the delta-flush hot path, bounded (24 entries / 7 days), cleared the moment the turn settles. Session resume folds the journaled tail back onto the restored transcript. When the backend also has a live text-only inflight projection for the same turn, the journal overlays its richer structure onto that row (longer text wins, base row id kept so live deltas keep landing) instead of treating it as caught up — the ordering defect that dropped locally recorded tool progress in the original PR. Co-authored-by: Omar Baradei <omar@kostudios.io> * fix(desktop): surface terminal error frames as failed bubbles message.complete frames with status "error" were detected only by a text regex heuristic, which misses the gateway's "Error: <detail>" texts and partial-text failures — a failed turn rendered as a healthy reply. The structured e…
randlee
pushed a commit
to randlee/hermes-agent
that referenced
this pull request
Aug 11, 2026
Follow-up cleanup for PR NousResearch#71123: - Extract getattr(args, 'lineage', 'single') == 'logical' to a local (appeared 3x in the export block) - Document that the double _collect_delegate_child_ids traversal in delete_session is an intentional TOCTOU guard inside the write txn
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What?
hermes sessions export --delete-after-verifiedpreviously exported and verified only the explicitly selected parent session.It then called
SessionDB.delete_session(), which intentionally cascade-deletes every delegate/subagent session associated with that parent. This meant delegate messages and transcript files could be permanently deleted even though their content was never included in the verified archive.Fix
The export flow now discovers the complete delegate deletion set and writes each affected session to a separate Markdown/QMD file.
Every generated file must pass the existing hash, session ID, and message-count verification before deletion begins. The expected deletion set is also checked inside the SQLite write transaction. If a new delegate appears while the export is running, deletion fails closed and all sessions are preserved.
Normal session deletion behavior remains unchanged.
Related Work
--delete-after-verifiedflow.These PRs introduced the affected workflow but do not handle delegate sessions that are cascade-deleted with their parent.
Tests
497 passedgit diff --checkpassed.