Based on #6079 by @tunamitom with critical fixes and comprehensive tests.
Changes from #6079:
- Fix: sanitization overwrite bug — Qwen message prep now runs AFTER codex
field sanitization, not before (was silently discarding Qwen transforms)
- Fix: missing try/except AuthError in runtime_provider.py — stale Qwen
credentials now fall through to next provider on auto-detect
- Fix: 'qwen' alias conflict — bare 'qwen' stays mapped to 'alibaba'
(DashScope); use 'qwen-portal' or 'qwen-cli' for the OAuth provider
- Fix: hardcoded ['coder-model'] replaced with live API fetch + curated
fallback list (qwen3-coder-plus, qwen3-coder)
- Fix: extract _is_qwen_portal() helper + _qwen_portal_headers() to replace
5 inline 'portal.qwen.ai' string checks and share headers between init
and credential swap
- Fix: add Qwen branch to _apply_client_headers_for_base_url for mid-session
credential swaps
- Fix: remove suspicious TypeError catch blocks around _prompt_provider_choice
- Fix: handle bare string items in content lists (were silently dropped)
- Fix: remove redundant dict() copies after deepcopy in message prep
- Revert: unrelated ai-gateway test mock removal and model_switch.py comment deletion
New tests (30 test functions):
- _qwen_cli_auth_path, _read_qwen_cli_tokens (success + 3 error paths)
- _save_qwen_cli_tokens (roundtrip, parent creation, permissions)
- _qwen_access_token_is_expiring (5 edge cases: fresh, expired, within skew,
None, non-numeric)
- _refresh_qwen_cli_tokens (success, preserve old refresh, 4 error paths,
default expires_in, disk persistence)
- resolve_qwen_runtime_credentials (fresh, auto-refresh, force-refresh,
missing token, env override)
- get_qwen_auth_status (logged in, not logged in)
- Runtime provider resolution (direct, pool entry, alias)
- _build_api_kwargs (metadata, vl_high_resolution_images, message formatting,
max_tokens suppression)
Fixes#6259. Three bugs fixed:
1. Config key mismatch: _get_client() and _start_daemon() read
'llmApiKey' (camelCase) but save_config() stores 'llm_api_key'
(snake_case). The config value was never read — only the env var
fallback worked.
2. Missing base URL support: users on OpenRouter or custom endpoints
had no way to configure HINDSIGHT_API_LLM_BASE_URL through setup.
Added llm_base_url to config schema with empty default, passed
conditionally to HindsightEmbedded constructor.
3. Daemon config change detection: config_changed now also checks
HINDSIGHT_API_LLM_BASE_URL, and the daemon profile .env includes
the base URL when set.
Keeps HINDSIGHT_API_LLM_API_KEY (with double API) in the daemon
profile .env — this matches the upstream hindsight .env.example
convention.
Documents the proxy env var support added in PR #3591 (salvage of #3411
by @kufufu9). Covers HTTPS_PROXY/HTTP_PROXY/ALL_PROXY precedence,
configuration methods, and scope.
* fix(tools): skip camofox auto-cleanup when managed persistence is enabled
When managed_persistence is enabled, cleanup_browser() was calling
camofox_close() which destroys the server-side browser context via
DELETE /sessions/{userId}, killing login sessions across cron runs.
Add camofox_soft_cleanup() — a public wrapper that drops only the
in-memory session entry when managed persistence is on, returning True.
When persistence is off it returns False so the caller falls back to
the full camofox_close(). The inactivity reaper still handles idle
resource cleanup.
Also surface a logger.warning() when _managed_persistence_enabled()
fails to load config, replacing a silent except-and-return-False.
Salvaged from #6182 by el-analista (Eduardo Perea Fernandez).
Added public API wrapper to avoid cross-module private imports,
and test coverage for both persistence paths.
Co-authored-by: Eduardo Perea Fernandez <el-analista@users.noreply.github.com>
* fix(doctor): only check the active memory provider, not all providers unconditionally
hermes doctor had hardcoded Honcho Memory and Mem0 Memory sections that
always ran regardless of the user's memory.provider config setting. After
the swappable memory provider update (#4623), users with leftover Honcho
config but no active provider saw false 'broken' errors.
Replaced both sections with a single Memory Provider section that reads
memory.provider from config.yaml and only checks the configured provider.
Users with no external provider see a green 'Built-in memory active' check.
Reported by community user michaelruiz001, confirmed by Eri (Honcho).
---------
Co-authored-by: Eduardo Perea Fernandez <el-analista@users.noreply.github.com>
Problem: hermes -w sessions accumulated 37+ worktrees and 1200+ orphaned
branches because:
- _cleanup_worktree bailed on any dirty working tree, but agent sessions
almost always leave untracked files/artifacts behind
- _prune_stale_worktrees had the same dirty-check, so stale worktrees
survived indefinitely
- pr-* and hermes/* branches from PR review had zero cleanup mechanism
Changes:
- _cleanup_worktree: check for unpushed commits instead of dirty state.
Agent work lives in pushed commits/PRs — dirty working tree without
unpushed commits is just artifacts, safe to remove.
- _prune_stale_worktrees: three-tier age system:
- Under 24h: skip (session may be active)
- 24h-72h: remove if no unpushed commits
- Over 72h: force remove regardless
- New _prune_orphaned_branches: on each -w startup, deletes local
hermes/hermes-* and pr-* branches with no corresponding worktree.
Protects main, checked-out branch, and active worktree branches.
Tests: 42 pass (6 new covering unpushed-commit logic, force-prune
tier, and orphaned branch cleanup).
- Fire on_session_finalize and on_session_reset in gateway _handle_reset_command()
- Fire on_session_finalize during gateway stop() for each active agent
- Move CLI test from tests/ root to tests/cli/ (matches recent restructure)
- Add 5 gateway tests covering reset hooks, ordering, shutdown, and error handling
- Place on_session_reset after new session is guaranteed to exist (covers
the get_or_create_session fallback path)
Plugins can now subscribe to session boundary events via
ctx.register_hook('on_session_finalize', ...) and
ctx.register_hook('on_session_reset', ...).
on_session_finalize — fires during CLI exit (/quit, Ctrl-C) and
before /new or /reset, giving plugins a chance to flush or clean up.
on_session_reset — fires after a new session is created via
/new or /reset, so plugins can initialize per-session state.
Closes#5592
Hermes Agent identified and patched its own prompting blind spots through
automated self-evaluation — running 64+ tool-use benchmarks across GPT-5.4
and Codex-5.3, diagnosing 5 failure modes, writing targeted prompt patches,
and verifying the fix in a closed loop.
Failure modes discovered and fixed:
- Mental arithmetic (wrong answers: 39,152,053 vs correct 39,151,253)
- User profile hallucination ('Windows 11' when running on Linux)
- Time guessing without verification
- Clarification-seeking instead of acting ('open where?' for port checks)
- Hash computation from memory (SHA-256, encodings)
- Confusing system RAM with agent's own persistent memory store
Two new XML sections added to OPENAI_MODEL_EXECUTION_GUIDANCE:
- <mandatory_tool_use>: explicit categories that must always use tools
- <act_dont_ask>: default to action on obvious interpretations
Results:
gpt-5.4: 68.8% → 100% tool compliance (+31.2pp)
gpt-5.3-codex: 62.5% → 100% tool compliance (+37.5pp)
Regression: 0/8 conversational prompts over-tooled
Anthropic signs thinking blocks against the full turn content. Any
upstream mutation (context compression, session truncation, orphan
stripping, message merging) invalidates the signature, causing HTTP 400
'Invalid signature in thinking block' — especially in long-lived
gateway sessions.
Strategy (following clawdbot/OpenClaw pattern):
1. Strip thinking/redacted_thinking from all assistant messages EXCEPT
the last one — preserves reasoning continuity on the current
tool-use chain while avoiding stale signature errors on older turns.
2. Downgrade unsigned thinking blocks to plain text — Anthropic can't
validate them, but the reasoning content is preserved.
3. Strip cache_control from thinking/redacted_thinking blocks to
prevent cache markers from interfering with signature validation.
4. Drop thinking blocks from the second message when merging
consecutive assistant messages (role alternation enforcement).
5. Error recovery: on HTTP 400 mentioning 'signature' and 'thinking',
strip all reasoning_details from the conversation and retry once.
This is the safety net for edge cases the proactive stripping
misses.
Addresses the issue reported in PR #6086 by @mingginwan while
preserving reasoning continuity (their PR stripped ALL thinking
blocks unconditionally).
Files changed:
- agent/anthropic_adapter.py: thinking block management in
convert_messages_to_anthropic (strip old turns, downgrade unsigned,
strip cache_control, merge-time strip)
- run_agent.py: one-shot signature error recovery in retry loop
- tests/test_anthropic_adapter.py: 10 new tests covering all cases
Gateway and cron had inconsistent reasoning_effort resolution:
- CLI: config.yaml only (correct)
- Gateway: config.yaml first, env var fallback
- Cron: env var first, config.yaml fallback
All three now read exclusively from agent.reasoning_effort in config.yaml.
Removed HERMES_REASONING_EFFORT env var support entirely — .env is for
secrets only, not behavioral config.
Previously crash recovery recreated detached sessions as if they were
fully managed, so polls and kills could lie about liveness and the
checkpoint could forget recovered jobs after the next restart.
This commit refreshes recovered host-backed sessions from real PID
state, keeps checkpoint data durable, and preserves notify watcher
metadata while treating sandbox-only PIDs as non-recoverable.
- Persist `pid_scope` in `tools/process_registry.py` and skip
recovering sandbox-backed entries without a host-visible PID handle
- Refresh detached sessions on access so `get`/`poll`/`wait` and active
session queries observe exited processes instead of hanging forever
- Allow recovered host PIDs to be terminated honestly and requeue
`notify_on_complete` watchers during checkpoint recovery
- Add regression tests for durable checkpoints, detached exit/kill
behavior, sandbox skip logic, and recovered notify watchers
Import hermes_constants.get_hermes_home at module level so it is
available in _start_daemon() when local mode starts the embedded
daemon. Previously the import was only inside _load_config(), causing
NameError when _start_daemon() referenced get_hermes_home().
Fixes#5993
Co-Authored-By: 史官 <historian@slock.team>
Add documentation for cocktailpeanut's hermes-mod community tool —
a web UI for creating and managing Hermes skins visually. Covers
installation (Pinokio, npx, manual), usage walkthrough, and feature
overview including ASCII art generation from images.
Ref: https://github.com/cocktailpeanut/hermes-mod
- DEFAULT_RESULT_SIZE_CHARS: 50K -> 100K (match current _LARGE_RESULT_CHARS)
- DEFAULT_PREVIEW_SIZE_CHARS: 2K -> 1.5K (match current _LARGE_RESULT_PREVIEW_CHARS)
- Per-tool overrides all set to 100K (terminal, execute_code, search_files)
- Remove pre-read byte guard (no behavioral regression vs current main)
- Revert limit signature change to int=500 (match current default)
- Restore original read_file schema description
- Update test assertions to match 100K thresholds
Add BudgetConfig dataclass to centralize and make overridable the
hardcoded constants (50K per-result, 200K per-turn, 2K preview) that
control when tool outputs get persisted to sandbox. Configurable at
the RL environment level via HermesAgentEnvConfig fields, threaded
through HermesAgentLoop to the storage layer.
Resolution: pinned (read_file=inf) > env config overrides > registry
per-tool > default. CLI override: --env.turn_budget_chars 80000
Simplify the vision auto-detection chain from 5 backends (openrouter,
nous, codex, anthropic, custom) down to 3:
1. OpenRouter (known vision-capable default model)
2. Nous Portal (known vision-capable default model)
3. Active provider + model (whatever the user is running)
4. Stop
This is simpler and more predictable. The active provider step uses
resolve_provider_client() which handles all provider types including
named custom providers (from #5978).
Removed the complex preferred-provider promotion logic and API-level
fallback — the chain is short enough that it doesn't need them.
Based on PR #5376 by Mibay. Closes#5366.
_deliver_result() now returns Optional[str] — None on success, error
message on failure. All failure paths (unknown platform, platform
disabled, config load error, send failure, unresolvable target)
return descriptive error strings.
mark_job_run() gains delivery_error param, tracked as
last_delivery_error on the job — separate from agent execution errors.
A job where the agent succeeded but delivery failed shows
last_status='ok' + last_delivery_error='...'.
The cronjob list tool now surfaces last_delivery_error so agents and
users can see when cron outputs aren't arriving.
Inspired by PR #5863 (oxngon) — reimplemented with proper wiring.
Tests: 3 new mark_job_run tests + 6 new _deliver_result return tests.
Add button-based exec approval to the Feishu adapter, matching the
existing Discord, Telegram, and Slack implementations.
When the agent encounters a dangerous command, Feishu users now see
an interactive card with four buttons instead of text instructions:
- Allow Once (primary)
- Allow Session
- Always Allow
- Deny (danger)
Implementation:
- send_exec_approval() sends an interactive card via the Feishu
message API with buttons carrying hermes_action in their value dict
- _handle_card_action_event() intercepts approval button clicks
before routing them as synthetic commands, directly calling
resolve_gateway_approval() to unblock the agent thread
- _update_approval_card() replaces the orange approval card with a
green (approved) or red (denied) status card showing who acted
- _approval_state dict tracks pending approval_id → session_key
mappings; cleaned up on resolution
The gateway's existing routing in _approval_notify_sync already checks
getattr(type(adapter), 'send_exec_approval', None) and will
automatically use the button-based flow for Feishu.
Tests: 16 new tests covering send, callback resolution, state
management, card updates, and non-interference with existing card
actions.
Salvaged fixes from community PRs:
- fix(model_switch): _read_auth_store → _load_auth_store + fix auth store
key lookup (was checking top-level dict instead of store['providers']).
OAuth providers now correctly detected in /model picker.
Cherry-picked from PR #5911 by Xule Lin (linxule).
- fix(ollama): pass num_ctx to override 2048 default context window.
Ollama defaults to 2048 context regardless of model capabilities. Now
auto-detects from /api/show metadata and injects num_ctx into every
request. Config override via model.ollama_num_ctx. Fixes#2708.
Cherry-picked from PR #5929 by kshitij (kshitijk4poor).
- fix(aux): normalize provider aliases for vision/auxiliary routing.
Adds _normalize_aux_provider() with 17 aliases (google→gemini,
claude→anthropic, glm→zai, etc). Fixes vision routing failure when
provider is set to 'google' instead of 'gemini'.
Cherry-picked from PR #5793 by e11i (Elizabeth1979).
- fix(aux): rewrite MiniMax /anthropic base URLs to /v1 for OpenAI SDK.
MiniMax's inference_base_url ends in /anthropic (Anthropic Messages API),
but auxiliary client uses OpenAI SDK which appends /chat/completions →
404 at /anthropic/chat/completions. Generic _to_openai_base_url() helper
rewrites terminal /anthropic to /v1 for OpenAI-compatible endpoint.
Inspired by PR #5786 by Lempkey.
Added debug logging to silent exception blocks across all fixes.
Co-authored-by: Hermes Agent <hermes@nousresearch.com>
Free-tier Nous Portal users were getting mimo-v2-omni (a multimodal
model) for all auxiliary tasks including compression, session search,
and web extraction. Now routes non-vision tasks to mimo-v2-pro (a
text model) which is better suited for those workloads.
- Added _NOUS_FREE_TIER_AUX_MODEL constant for text auxiliary tasks
- _try_nous() accepts vision=False param to select the right model
- Vision path (_resolve_strict_vision_backend) passes vision=True
- All other callers default to vision=False → mimo-v2-pro
Currently, MCP servers are included on all platforms by default. If a
platform's toolset list does not explicitly name any MCP servers, every
globally enabled MCP server is injected. There is no way to opt a
platform out of MCP servers entirely.
This matters for the API server platform when used as an execution
backend — each spawned agent session gets the full MCP tool schema
injected into its system prompt, dramatically inflating token usage
(e.g. 57K tokens vs 9K without MCP tools) and slowing response times.
Add a "no_mcp" sentinel value for platform_toolsets. When present in a
platform's toolset list, all MCP servers are excluded for that platform.
Other platforms are unaffected.
Usage in config.yaml:
platform_toolsets:
api_server:
- terminal
- file
- web
- no_mcp # exclude all MCP servers
The sentinel is filtered out of the final toolset — it does not appear
as an actual toolset name.
- The MCP SDK Pydantic model uses camelCase (structuredContent), not
snake_case (structured_content). The original getattr was a silent no-op.
- When structuredContent is present, return it AS the result instead of
alongside text — the structured payload is the machine-readable data.
- Move test file to tests/tools/ and fix fake class to use camelCase.
- Patch _run_on_mcp_loop in tests so the handler actually executes.
MCP CallToolResult may include structured_content (a JSON object) alongside
content blocks. The tool handler previously only forwarded concatenated text
from content blocks, silently dropping the structured payload.
This breaks MCP tools that return a minimal human text in content while
putting the actual machine-usable payload in structured_content.
Now, when structured_content is present, it is included in the returned
JSON under the 'structuredContent' key.
FixesNousResearch/hermes-agent#5874
* fix(telegram): replace substring caption check with exact line-by-line match
Captions in photo bursts and media group albums were silently dropped when
a shorter caption happened to be a substring of an existing one (e.g.
"Meeting" lost inside "Meeting agenda"). Extract a shared _merge_caption
static helper that splits on "\n\n" and uses exact match with whitespace
normalisation, then use it in both _enqueue_photo_event and
_queue_media_group_event.
Adds 13 unit tests covering the fixed bug scenarios.
Cherry-picked from PR #2671 by Dilee.
* fix: extend caption substring fix to all platforms
Move _merge_caption helper from TelegramAdapter to BasePlatformAdapter
so all adapters inherit it. Fix the same substring-containment bug in:
- gateway/platforms/base.py (photo burst merging)
- gateway/run.py (priority photo follow-up merging)
- gateway/platforms/feishu.py (media batch merging)
The original fix only covered telegram.py. The same bug existed in base.py
and run.py (pure substring check) and feishu.py (list membership without
whitespace normalization).
* fix(auxiliary): resolve named custom providers and 'main' alias in auxiliary routing
Two bugs caused auxiliary tasks (vision, compression, etc.) to fail when
using named custom providers defined in config.yaml:
1. 'provider: main' was hardcoded to 'custom', which only checks legacy
OPENAI_BASE_URL env vars. Now reads _read_main_provider() to resolve
to the actual provider (e.g., 'custom:beans', 'openrouter', 'deepseek').
2. Named custom provider names (e.g., 'beans') fell through to
PROVIDER_REGISTRY which doesn't know about config.yaml entries.
Now checks _get_named_custom_provider() before the registry fallback.
Fixes both resolve_provider_client() and _normalize_vision_provider()
so the fix covers all auxiliary tasks (vision, compression, web_extract,
session_search, etc.).
Adds 13 unit tests. Reported by Laura via Discord.
---------
Co-authored-by: Dilee <uzmpsk.dilekakbas@gmail.com>
* fix(cli): route error messages through ChatConsole inside patch_stdout
Cherry-pick of PR #5798 by @icn5381.
Replace self.console.print() with ChatConsole().print() for 11 error/status
messages reachable during the interactive session. Inside patch_stdout,
self.console (plain Rich Console) writes raw ANSI escapes that StdoutProxy
mangles into garbled text. ChatConsole uses prompt_toolkit's native
print_formatted_text which renders correctly.
Same class of bug as #2262 — that fix covered agent output but missed
these error paths in _ensure_runtime_credentials, _init_agent, quick
commands, skill loading, and plan mode.
* fix(model-picker): add scrolling viewport to curses provider menu
Cherry-pick of PR #5790 by @Lempkey. Fixes#5755.
_curses_prompt_choice rendered items starting unconditionally from index 0
with no scroll offset. The 'More providers' submenu has 13 entries. On
terminals shorter than ~16 rows, items past the fold were never drawn.
When UP-arrow wrapped cursor from 0 to the last item (Cancel, index 12),
the highlight rendered off-screen — appearing as if only Cancel existed.
Adds scroll_offset tracking that adjusts each frame to keep the cursor
inside the visible window.
* feat(cli): skin-aware compact banner + git state in startup banner
Combined salvage of PR #5922 by @ASRagab and PR #5877 by @xinbenlv.
Compact banner changes (from #5922):
- Read active skin colors and branding instead of hardcoding gold/NOUS HERMES
- Default skin preserves backward-compatible legacy branding
- Non-default skins use their own agent_name and colors
Git state in banner (from #5877):
- New format_banner_version_label() shows upstream/local git hashes
- Full banner title now includes git state (upstream hash, carried commits)
- Compact banner line2 shows the version label with git state
- Widen compact banner max width from 64 to 88 to fit version info
Both the full Rich banner and compact fallback are now skin-aware
and show git state.
- Add 7 unit tests for _profile_arg: default home, named profile,
hash path, nested path, invalid name, systemd integration, launchd integration
- Add stt.local.language to config.yaml (empty = auto-detect)
- Both STT code paths now read config.yaml first, env var fallback,
then default (auto-detect for faster-whisper, 'en' for CLI command)
- HERMES_LOCAL_STT_LANGUAGE env var still works as backward-compat fallback
- gateway/run.py: Strip "(The user sent a message with no text content)"
placeholder when voice transcription succeeds — it was being appended
alongside the transcript, creating duplicate user turns.
- tools/transcription_tools.py: Wire HERMES_LOCAL_STT_LANGUAGE env var
into the faster-whisper backend. It was only used by the CLI fallback
path (_transcribe_local_command), not the primary faster-whisper path.
When a Discord voice message arrives, the adapter sets event.text to
"(The user sent a message with no text content)" since voice messages
have no text content. The transcription enrichment in
_enrich_message_with_transcription() then prepends the transcript but
leaves the placeholder intact, causing the agent to receive both:
[The user sent a voice message~ Here's what they said: "..."]
(The user sent a message with no text content)
The agent sees this as two separate user turns — one transcribed
and one empty — creating confusing duplicate messages.
Fix: when the transcription succeeds and user_text is only the empty
placeholder, return just the transcript without the redundant placeholder.
generate_launchd_plist() and generate_systemd_unit() were missing the
--profile <name> argument in ProgramArguments/ExecStart, causing
hermes gateway start to regenerate plists that fell back to
~/.hermes/active_profile instead of the intended profile.
Fix:
- Add _profile_arg(hermes_home?) helper returning '--profile <name>'
only for ~/.hermes/profiles/<name> paths, empty string otherwise.
- Update generate_launchd_plist() to build ProgramArguments array
dynamically with --profile when applicable.
- Update generate_systemd_unit() both user and system service
branches with {profile_arg} in ExecStart.
This ensures hermes --profile <name> gateway start produces a
service definition that correctly scopes to the named profile.
Follow-up to sroecker's PR #5918 — test mocks were using the old 3-arg
callback signature (name, preview, args) instead of the new
(event_type, name, preview, args, **kwargs).
When /v1/runs receives an OpenAI-style array of messages as input, all
messages except the last user turn are now extracted as conversation_history.
Previously only the last message was kept, silently discarding earlier
context in multi-turn conversations.
Handles multi-part content blocks by flattening text portions. Only fires
when no explicit conversation_history was provided.
Based on PR #5837 by pradeep7127.
Allow clients to pass explicit conversation_history in /v1/responses and
/v1/runs request bodies instead of relying on server-side response chaining
via previous_response_id. Solves problems with stateless deployments where
the in-memory ResponseStore is lost on restart.
Adds input validation (must be array of {role, content} objects) and clear
precedence: explicit conversation_history > previous_response_id.
Based on PR #5805 by VanBladee, with added input validation.
Commit cc2b56b2 changed the tool_progress_callback signature from
(name, preview, args) to (event_type, name, preview, args, **kwargs)
but the API server's chat completion streaming callback was not updated.
This caused tool calls to not display in Open WebUI because the
callback received arguments in wrong positions.
- Update _on_tool_progress to use new 4-arg signature
- Add event_type filter to only show tool.started events
- Add **kwargs for optional duration/is_error parameters
Telegram's set_message_reaction replaces all reactions in one call,
so _remove_reaction was never called (unlike Discord's additive model).
Default reactions to disabled — users opt in via telegram.reactions: true.
Mirror the Discord reaction pattern for Telegram:
- 👀 (eyes) when message processing begins
- ✅ (check) on successful completion
- ❌ (cross) on failure
Controlled via TELEGRAM_REACTIONS env var or telegram.reactions
in config.yaml (enabled by default, like Discord).
Uses python-telegram-bot's Bot.set_message_reaction() API.
Failures are caught and logged at debug level so they never
break message processing.
- ignored_channels: channels where bot never responds (even when mentioned)
- no_thread_channels: channels where bot responds directly without thread
Both support config.yaml and env vars (DISCORD_IGNORED_CHANNELS,
DISCORD_NO_THREAD_CHANNELS), following existing pattern for
free_response_channels.
Fixes#5881
The response validation stage unconditionally marked Codex Responses API
replies as invalid when response.output was empty, triggering unnecessary
retries and fallback chains. However, _normalize_codex_response can
recover from this state by synthesizing output from response.output_text.
Now the validation stage checks for output_text before marking the
response invalid, matching the normalization logic. Also fixes
logging.warning → logger.warning for consistency with the rest of the
file.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace 10 callsites across 6 files that manually opened config.yaml,
called yaml.safe_load(), and handled missing-file/parse-error fallbacks
with the new read_raw_config() helper from hermes_cli/config.py.
Each migrated site previously had 5-8 lines of boilerplate:
config_path = get_hermes_home() / 'config.yaml'
if config_path.exists():
import yaml
with open(config_path) as f:
cfg = yaml.safe_load(f) or {}
Now reduced to:
from hermes_cli.config import read_raw_config
cfg = read_raw_config()
Migrated files:
- tools/browser_tool.py (4 sites): command_timeout, cloud_provider,
allow_private_urls, record_sessions
- tools/env_passthrough.py: terminal.env_passthrough
- tools/credential_files.py: terminal.credential_files
- tools/transcription_tools.py: stt.model
- hermes_cli/commands.py: config-gated command resolution
- hermes_cli/auth.py (2 sites): model config read + provider reset
Skipped (intentionally):
- gateway/run.py: 10+ sites with local aliases, critical path
- hermes_cli/profiles.py: profile-specific config path
- hermes_cli/doctor.py: reads raw then writes fixes back
- agent/model_metadata.py: different file (context_length_cache.yaml)
- tools/website_policy.py: custom config_path param + error types
