Absorbs contributor PR #3544 (@rodboev, closes#3340) with two fixes:
1. DETECTION VOCAB (would never fire): the original gated on action names
{save,create,update,upsert}, which don't match the real agent tool enums —
memory.action is add|replace|remove, skill_manage.action is
create|patch|edit|delete|write_file|remove_file. Split into per-tool
predicates with the correct vocabularies: _isMemorySave gates memory on
{add,replace}; _isSkillUpdate gates skill_manage on {create,patch,edit,
write_file}. Deletions excluded so the saved/updated verbs stay accurate;
running/errored excluded.
2. SNAPSHOT/RESTORE PERSISTENCE (Codex catch): classification lived only on the
row._tcData JS property, which does NOT survive the outerHTML/innerHTML
snapshot+restore the live tool-call group uses on session switch/restore —
a restored memory/skill row would be re-counted as a generic tool and the
suffix would silently vanish. buildToolCard now also stamps durable
data-memory-save / data-skill-update attributes, and _syncToolCallGroupSummary
counts them as a fallback when _tcData is absent. Verified live across a real
outerHTML round-trip: label identical before/after.
Replaces the PR's static source assertions with a node-driven behavioral test
(11 cases) covering the real action vocabularies, exclusions, case-insensitivity,
null-arg safety, and the durable-attribute persistence guard.
Co-authored-by: rodboev <rodboev@users.noreply.github.com>
Adds the Verdigris dark-only appearance skin (emerald/forest-green + bronze-gold),
renamed from the contributor's 'Hermes Agent' to a descriptive material name per
maintainer naming convention. Registered across all 5 skin sites (config allowlist,
boot.js swatch, index.html FOUC map, i18n in 12 locales, scoped CSS palette) + test.
Also fixes the zeus i18n test (zeus is no longer the trailing skin token).
Co-authored-by: rodboev <rodboev@users.noreply.github.com>
Closes#2771.
v0.51.117 (PR #2766) introduced a top-level function _inflightStateLimits()
in static/ui.js that collided with the window._inflightStateLimits config
object set in static/boot.js. Because top-level function declarations in
classic (non-module) scripts attach to window, boot.js's assignment
overwrote the function reference, and every later _inflightStateLimits()
call threw TypeError. _compactInflightState() runs on every send(), so
no new chat session could be created — v0.51.117 is effectively unusable.
Reported by @jahilldev, with multiple users (@isma3iloiso, @theDanielJLewis,
@JHVenn) confirming the bug or reverting to v0.51.116.
Fix: rename the function to _getInflightStateLimits() — the window-attached
config key stays under its original name (unchanged for any downstream
code that reads it). Updates all 4 call sites in static/ui.js.
Tests:
- Update tests/test_inflight_storage_quota.py — the existing test
asserted 'function _inflightStateLimits()' in UI_JS as a positive
presence check, which certified the bug. Now asserts the renamed
function name is present AND the old colliding name is absent AND
no stale call sites remain.
- Add tests/test_window_function_collision.py — generalized regression
that scans every static JS file for top-level function declarations
whose name also appears as the target of 'window.X = {...}' or
'window.X = <number>'. This is the exact shape that broke #2715
(_pinnedSessionsLimit in v0.51.106) and #2771. Test fails with a
precise diagnostic naming the file and symbol if the bug class
returns. Confirmed test FAILS on current master (unfixed) and PASSES
on this branch.
Verified end-to-end against the live browser before commit:
- typeof window._inflightStateLimits === 'object' (config preserved)
- typeof window._getInflightStateLimits === 'function'
- _getInflightStateLimits() returns the limits object
- saveInflightState() persists to localStorage without throwing
Full pytest suite: 6308 passed, 6 skipped, 3 xpassed, 8 subtests passed.
Opus advisor: SHIP.
Without --force, git fetch origin --tags refuses to overwrite divergent
local tags and returns 'would clobber existing tag', jamming the entire
WebUI update path indefinitely. The WebUI is a release-tracking consumer
that never pushes tags, so it should always defer to whatever the remote
says a release tag points to. Add --force to all three fetch-tag call
sites:
- _check_repo (the 'Check now' button + periodic check)
- apply_force_update (force-reset to remote HEAD)
- apply_update (stash + pull --ff-only)
Tests:
- Updated 3 existing tests in test_updates.py whose fake_git mocks
asserted the exact ['fetch', 'origin', '--tags'] args list.
- Updated 1 existing test in test_update_banner_fixes.py that asserted
the same shape for apply_update.
- Added 4 new regression tests:
- test_check_repo_fetches_tags_with_force
- test_apply_force_update_fetches_tags_with_force
- test_apply_update_fetches_tags_with_force
- test_check_repo_recovers_from_remote_retag (end-to-end,
proves the bare --tags fetch shape is no longer used)
Closes#2756.
Opus advisor review of stage-375 flagged that the protected-bracket set including `<` and `>` caused tables containing comparison operators across adjacent columns to mis-collapse: `| x < 5 | y > 10 |` matched `< ... >` as a bracket pair and stashed the inner pipe, producing one cell instead of two.
Real LLM table output uses angle brackets as comparison operators far more often than as content-grouping pairs, so the safer default is to NOT treat them as a matched pair. Dropped `<` from the opener class and `>` from both closer classes.
Three regression tests added (`TestComparisonOperatorsAcrossColumns` class): `< … >` across columns, `<` alone, `>` alone.
PR #2428's iterative _protectPipes regex introduced two issues we caught during stage assembly:
1. The negated character classes [^)\]}'>] added `'` as a stop character. That breaks cells containing string-literal pipes like `('a'|'b')` (Python type-union examples) — they would still mis-split. Dropped the apostrophe-stop.
2. The literal `}` inside the regex character classes confused the brace-counting extractFunc driver in tests/test_renderer_js_behaviour.py, breaking all 45 existing node-driven renderer tests. Rewrote both brace literals as hex escapes (\\x7b and \\x7d) — semantically identical at the regex-engine level but the JS source carries no bare brace glyph.
Also added tests/test_issue2428_table_pipe_protection.py with 9 regression tests covering single-pipe, multi-pipe-in-brackets, apostrophes-with-pipes, and the KaTeX \$...\$ guard.
Opus advisor on stage-371 caught three issues during pre-release review:
1. RTL salvage missed KaTeX math (display equations + inline LaTeX), diff
blocks, CSV tables (column order must read left-to-right regardless of
chat direction), and .skill-file-path. The first salvage commit only
covered pre/code/kbd/samp/tt and tool-call bodies. Added a second
force-LTR block covering: .katex, .katex-block, .katex-display,
.katex-html, .katex-inline, .diff-block (+children), .csv-table-wrap,
.csv-table (+children), .skill-file-path. Severity: KaTeX is the most
user-visible gap — any user rendering math under RTL would see flipped
equations.
2. Quota chip @media (max-width:1400px) hide rule conflicted at exactly
1400px with the existing @media (min-width:1400px) .messages-inner
rule — chip was hidden AT the wide-desktop boundary where it should
first appear. Changed to (max-width:1399.98px). Visually verified at
1400px: chip now correctly visible there.
3. Dead .icon-btn.provider-quota-chip selector — chip never has icon-btn
class. Removed.
Test added: test_rtl_math_and_tables_stay_ltr (pins the 4 new LTR
surfaces). Also removed dead code in test_rtl_code_blocks_stay_ltr
(unused code_block variable).
Per stage-fix protocol: SHIP-with-followup applied on the stage rather
than the source PR, since #2409 is already merged-into-stage and
nesquena-approved. Stage-371 review-bypass batch path still holds.
Move the `_PORTAL_PROVIDERS` guard in `resolve_model_provider()` to run
BEFORE the `prefix == config_provider` strip branch. The guard was added
for NVIDIA (along with the Nous portal cases in #854 / #894) but was
placed after the strip, so it never fired when `config_provider == "nvidia"`
and the model id started with `nvidia/`.
For `model_id="nvidia/nemotron-3-super-120b-a12b"`,
`config_provider="nvidia"`:
- prefix = "nvidia", bare = "nemotron-3-super-120b-a12b"
- prefix == config_provider → True → strip branch returned bare name
- `_PORTAL_PROVIDERS` guard never reached
- bare "nemotron-3-super-120b-a12b" sent to NVIDIA NIM → HTTP 404
NIM requires the full namespaced path. The fix moves the portal guard
to run first, so all portal providers (Nous, OpenCode-Zen, OpenCode-Go,
NVIDIA NIM) always preserve the full `provider/model` id regardless of
whether the prefix happens to equal the provider name.
This also closes a latent symmetric bug for the Nous case if a
`nous/<model>` id ever existed in the catalog.
Test plan:
- New `tests/test_issue2177_nvidia_prefix_preservation.py` covers:
- nvidia/nemotron-... under nvidia (the reported case)
- cross-namespace qwen/ and meta/ under nvidia (regression pin)
- every static nvidia model in `_PROVIDER_MODELS` resolves to itself
- latent nous/<model> under nous (structural ordering pin)
- non-portal providers (anthropic) still strip — fix doesn't over-correct
- Existing portal-routing suites (test_nous_portal_routing.py,
test_issue895_894_nous_prefix.py) continue to pass.
- Full test suite: 5320 passed, 4 skipped, 3 xpassed.
Reported on Discord by @vishnu (Nathan forwarded as #2177).
Opus advisor pass on stage-341 found three surgical items:
1. static/i18n.js:it — PR #2064 branched before stage-340 landed the 'it'
locale (#2067), missing 9 session_*worktree* keys. Mechanical mirror of
en/ja position. Italian falls back to English silently without this fix.
2. api/streaming.py — PR #2107's new break short-circuit was silent in both
the aux and agent title-generation paths. Added logger.debug calls before
each break so production logs surface the exit shape.
3. api/streaming.py — Expanded _title_should_skip_remaining_attempts docstring
to document the membership criterion explicitly (vs the implicit
reasoning-only-burn case it ships with today). Future additions
(llm_safety_blocked, llm_oauth_quota) have a clear inclusion test.
CHANGELOG updated under the Stage-341 maintainer fixes section to mirror
the stage-340 pattern. All targeted tests pass (57/57 in the affected
modules).
Renames the [Unreleased] section to [v0.51.47] (Release W, shipped today
via stage-340) and folds in the stage-341 batch — PR #2105 RFC, PR #2107
title-retry fix, PR #2064 worktree archive copy, plus the stage-341
maintainer fix (RFC conventions guidance).
Also removes the duplicate v0.51.46 heading line that landed in v0.51.47's
stage-340 merge (the duplicate was a no-op — empty body line under the
extra heading — but tidying it up here.
When merging PR #2105 (Hermes Run Adapter RFC) the standing concern was
that landing the RFC unconfirmed would invite the speculative-fragment
implementation pattern we just had to put on hold with PR #2071 — well-
written 651-LOC standalone scripts with no callers.
Add a single bullet to the conventions block so the contract is explicit:
an RFC is a design direction, not an invitation to PR fragments against
it. Implementation slices need maintainer confirmation first.
Applied during stage-341 build, not requested from @Michaelyklam — the
guardrail belongs in the conventions doc itself rather than as a one-off
ask on this PR.
Reasoning models (Qwen3-thinking via LM Studio, DeepSeek-R1, Kimi-K2,
etc.) can burn their entire output budget on hidden reasoning tokens and
emit no visible content. The previous title-generation retry path
classified that as llm_length and doubled the budget — but the second
call produces the same shape, so the retry only doubled the GPU/credit
burn. Repeated across the two prompts in _title_prompts() this came to
~3000 reasoning tokens of GPU work per new chat. On local LM Studio
servers behind a custom: provider (where is_lmstudio=False means
reasoning_effort: none never reaches the model) it manifested as the GPU
never going idle after a prompt.
Fix:
- _extract_title_response: classify reasoning-bearing empty responses
as llm_empty_reasoning regardless of finish_reason. The presence of
reasoning_content is the diagnostic signal, not finish_reason.
- _title_retry_status: drop llm_empty_reasoning from the retry set.
Length-truncated responses WITHOUT reasoning still retry (those are
legitimately recoverable by a larger budget).
- Add _title_should_skip_remaining_attempts() and break out of the
prompt-iteration loop on empty-reasoning. A second prompt against
the same model would produce the same shape.
- Falls through to _fallback_title_from_exchange for a local-summary
title.
Tests updated to invert the previous reasoning-retry assertions:
- test_aux_short_circuits_on_empty_reasoning_without_retrying
- test_aux_still_retries_finish_length_without_reasoning
- test_agent_route_short_circuits_on_empty_reasoning_without_retrying
- test_agent_route_still_retries_finish_length_without_reasoning
Companion agent-side work (LM Studio classifier for custom: providers)
is tracked separately on the hermes-agent side; this WebUI fix is the
belt-and-braces guard so the loop stops regardless of agent classifier
state.
Reported by @darkopetrovic. Closes#2083.
Co-authored-by: darkopetrovic <darkopetrovic@users.noreply.github.com>
(cherry picked from commit efeae4a86e377069c0f09d140429ecb111a8dd1a)
Opus SHOULD-FIX from stage-340 review. PR #2067 added the it locale
between en and ja; PR #2100 added 4 toast keys to 8 other locales but
missed it. Falls back to English via t() defaults so no user-visible
break, but it's an i18n parity hole.
4 LOC, mechanical add inside the it: block at the canonical position
(immediately after cron_profile_server_default_hint, mirroring en/ja).
Co-authored-by: ai-ag2026 <261867348+ai-ag2026@users.noreply.github.com>
Co-authored-by: Samuel Gudi <samuel.gudi.official@gmail.com>
PR #2067 made TestVoiceModePreferenceGate.test_settings_pane_has_voice_mode_i18n_keys
adaptive via self.LOCALES but only defined LOCALES on the sibling class
TestComposerVoiceButtonI18n. AttributeError on CI.
Mirror the tuple to TestVoiceModePreferenceGate so the count assert resolves
to 10 with Italian present.
Co-authored-by: Samuel Gudi <samuel.gudi.official@gmail.com>
Opus stage-339 review SHOULD-FIX items:
1. server.py: drop 'unsafe-eval' from CSP report-only policy.
Verified by grepping all production JS — zero matches for eval(),
new Function(), or string-form setTimeout/setInterval. Keeping it
was a gratuitous privilege.
2. server.py: add https://cdn.jsdelivr.net to script-src + style-src.
index.html loads Prism/xterm/katex from this CDN with SRI hashes —
without the allowance every page load fires known-good CSP violations
that drown out real signal once a collector is wired.
3. api/commands.py: sanitize plugin command error. Previously returned
f'Plugin command error: {exc}' which would leak paths/env from
FileNotFoundError('/etc/something/secret.key') etc. Now returns only
the exception type name; full traceback goes to server log.
Test asserts updated to match the new policy shape.
Co-authored-by: Opus advisor <opus-advisor@hermes.local>
Opus stage-338 review SHOULD-FIX: silent drop at api/providers.py:1049
was diagnostically opaque. logger.warning() now surfaces the bad
config entry so operators can spot misconfigurations.
Co-authored-by: Opus advisor <opus-advisor@hermes.local>
Update CONTRIBUTORS.md and the README contributors section to reflect
130 contributors and 568 PR credits as of v0.51.44 (was 66/142 at
v0.50.245). The numbers grew because:
- The previous refresh was 1 release-cycle ago (50+ tags + 8 batch
releases of contributor PRs ago).
- The new counting rule explicitly includes closed-but-absorbed PRs:
PRs whose original branch shows "closed" on GitHub but whose content
shipped via batch-release squash with a Co-authored-by trailer, or
via salvage rewrite with CHANGELOG attribution. This better reflects
what users actually contributed.
The compilation pipeline:
1. Pull every closed PR from gh api (state=closed, both merged and
unmerged on GitHub) — 1421 PRs.
2. Walk CHANGELOG.md release-by-release and extract:
- `PR #N by @user` (canonical bullet form)
- `(#N by @user`, `(PR #N by @user`, `(#N, @user;`
- `PRs #A, #B by @user` (plural)
- `@user — PR #N`, `@user — N PR (#A, #B)`
- `(credit: @user)` and `(credit: @userA and @userB)`
3. For every PR# mentioned in CHANGELOG, union the explicit @-attributed
users with the gh PR author (when external). Maintainer accounts
(@nesquena, @nesquena-hermes) are excluded.
4. For PRs merged on GitHub but not mentioned in CHANGELOG (very early
PRs, non-noteworthy direct merges), credit the gh author.
5. Three salvaged-design contributors not directly in CHANGELOG are
credited in the special-thanks roll: @indigokarasu (#213 →
v0.50.0 design language), @andrewy-wizard (#177 → initial Chinese
locale absorbed into v0.42.0), @zenc-cp (#133 → anti-hallucination
guard absorbed into streaming.py).
Pre-cleaning step strips HTML entities (` ` etc.) before PR# scan
to avoid false matches. PR# regex requires a whitespace/paren/bracket
preceder so identifiers like `--key=123` and `(##10`-style headings
don't pollute the count.
Per-user first/last release computed from:
- For merged-on-GH PRs: the smallest tag whose creator-date is >= the
PR's merged_at timestamp.
- For absorbed PRs: the release section in CHANGELOG that explicitly
attributes to the user (or the earliest release that mentions the
PR# if no explicit attribution exists for that user).
CONTRIBUTORS.md sections:
- Top contributors (5+ PRs) — 20 people, ranked
- Sustained contributors (3–4 PRs) — 11 people
- Two-PR contributors — 14 people, flat list
- Single-PR contributors — 85 people, flat list
- How credit is tracked — four paths described
- Special thanks — 11 highlight blurbs
README contributors section trimmed to top-10 table + notable-
contribution blurbs (29 distinct contributors mentioned with concrete
PR numbers). Same data, condensed for the README.
No code changes. Docs only.
CI's pytest invocation imports conftest twice (once via the standard
tests/ discovery, once via repo-root rootdir discovery), producing two
distinct function objects with the same __qualname__ but different `is`
identity. The strict identity assertion failed because each import
created a fresh closure. Switch to __qualname__ substring check — same
guarantee (default-on state has the wrapper installed; fixture restores
the real one) without the multi-import sensitivity.
CI on Python 3.11 still failed test_allow_outbound_network_fixture_*
because the previous module-global toggle (_ALLOW_OUTBOUND=True/False)
was unreliable on the runner — the wrapper's global lookup at call time
sometimes saw False even after the fixture's True assignment.
Switch to monkeypatch-based fixture: instead of toggling a global that
the wrapper checks, restore socket.create_connection and
socket.socket.connect to their REAL captured implementations for the
duration of the test. Pytest's monkeypatch fixture handles teardown so
the wrappers are reinstalled automatically.
Rewrote the two paired tests to check function identity
(socket.create_connection is _hermes_blocked_create_connection vs. is
_REAL_CREATE_CONNECTION) instead of attempting a live outbound to
8.8.8.8:53 — direct identity check is hermetic and doesn't depend on
whether the CI runner has any outbound network access at all.
Two low-severity follow-ups from Opus regrounding review:
1. The IPv6 unique-local fc00::/7 check was `h.startswith('fc') or
h.startswith('fd')` — too loose. It would also classify hostnames
like 'food.example.com' or 'fdsa.test' as 'local' and silently let
them through the block. Tightened to a regex match for canonical
IPv6 syntax (`f[cd][0-9a-f]{0,2}:`) so only actual IPv6 addresses
match. Same fix in both tests/conftest.py and server.py.
2. test_allow_outbound_network_fixture_unblocks was technically
self-passing: it tried to connect to a *.invalid hostname, which is
in the allow-list, so the real socket.create_connection would run
regardless of whether the fixture toggled the block. Replaced with
a public-IP-based test that actually proves the toggle works, plus
a paired test_block_is_active_outside_the_fixture sanity test that
proves the block is on without the fixture.
Both follow-ups noted by Opus advisor as 'defer-OK' but trivial fixes
so landing them in this batch.
CI on Python 3.13 (clean editable install, no hermes_cli package) was still
failing the 3 lmstudio tests after the first fix attempt. Root cause: the
outer try/except in the lmstudio branch was catching ImportError from
`from hermes_cli.models import provider_model_ids`, hijacking the whole
branch and silently skipping the urlopen fallback.
Restructured into two independent tiers:
1. hermes_cli lookup in its own try/except — ImportError logs at DEBUG
and continues with lm_ids=[].
2. urlopen fallback runs unconditionally when lm_ids is empty, including
after hermes_cli import failure.
New regression test `test_lmstudio_fallback_works_when_hermes_cli_unavailable`
explicitly blocks hermes_cli via sys.meta_path and verifies the lmstudio
group still populates from the urlopen fallback. Without this test, the
CI-vs-local divergence (local env had hermes_cli installed, CI didn't)
would keep slipping through.
All 12 lmstudio-related tests pass, including the 3 #1527 tests that
broke on stage-337.
PR #2053 added worktree-backed session creation. PR #2041 (shipped in
v0.51.42) added state.db sidecar reconciliation that rebuilds a missing
<sid>.json sidecar from the canonical state.db row when the JSON file is
gone (failed save, manual rm, restore-from-backup with mismatched dirs).
The two interact silently. `_state_db_row_to_sidecar()` was hard-coding
`'workspace': ''` and never propagating the four worktree_* fields from
the row to the rebuilt sidecar dict. So a worktree-backed session that
loses its sidecar and gets rebuilt from state.db:
- loses `worktree_path` → matches the empty-session sidebar filter at
`api/models.py:1067/1107` (which spares worktree-backed empty sessions
via `not s.get('worktree_path')`) → session disappears from the
sidebar even though the worktree directory still exists on disk.
- loses `workspace` → downstream tools (terminal panels, file pickers
that use `s.workspace`) operate on empty string instead of the original
worktree path.
- always reports `message_count == 0` → contributes to the empty-session
filter even for sessions that have messages in `state.db.messages`.
Fix:
1. `_read_state_db_missing_sidecar_rows()` SELECT now includes
`workspace, worktree_path, worktree_branch, worktree_repo_root,
worktree_created_at, message_count` (each gated by
`_sql_optional_col()` so older state.db schemas without those columns
continue to work — recovery degrades gracefully rather than 500ing).
2. `_state_db_row_to_sidecar()` propagates each field. workspace comes
from the row if it's a string, otherwise '' (matching pre-fix behavior
for non-worktree sessions). message_count comes from the row if
it's an int, otherwise falls back to `len(messages)` so the rebuilt
sidecar always has a coherent count.
3 new regression tests in tests/test_state_db_worktree_recovery.py
exercise:
- worktree session with messages → all four worktree_* fields preserved.
- non-worktree session → worktree_* fields all None (no spurious
propagation), workspace=''.
- empty worktree session (the worst case) → confirms the rebuilt sidecar
does NOT match the empty-session-exempt filter, so it stays visible
in the sidebar.
Caught by Opus advisor during stage-337 review (the cross-PR interaction
between #2053 and the previously-shipped #2041 wasn't exercised by either
PR's individual test suite).
PR #1970 added a dedicated `elif pid == "lmstudio":` branch in
`get_available_models()` that fetches the live /v1/models list when the
hermes_cli helper doesn't have ids cached. The fallback path inside that
branch only looked at `cfg["providers"]["lmstudio"]["base_url"]`, missing
the historical config shape where the URL lives under `cfg["model"]`:
model:
provider: lmstudio
base_url: http://192.168.1.22:1234/v1 ← here, not under providers.lmstudio
providers:
lmstudio:
api_key: local-key
3 pre-existing tests in tests/test_issue1527_lmstudio_base_url_classification
broke on stage-337 because of this — they passed on master, failed after
the PR #1970 merge.
The simpler fix is to enhance the already-introduced `_get_provider_base_url()`
helper so it falls back to `cfg["model"]["base_url"]` when
`cfg["model"]["provider"] == provider_id`, then use the helper inside the
lmstudio branch instead of a direct lookup. This keeps the previous
behaviour (where the generic configured-provider branch handled lmstudio
via the model block) while preserving PR #1970's live-discovery additions.
Belt-and-suspenders: `_get_provider_base_url()` explicitly does NOT inherit
model.base_url for providers other than the active one — if a user's config
says `model.provider: anthropic` and they have `providers.openai` configured
without a base_url, openai must still resolve to None (use SDK default),
not to the anthropic proxy URL.
6 new regression tests in tests/test_pr1970_lmstudio_base_url_fallback.py
lock the two-location lookup, the precedence rule (explicit providers entry
wins over model fallback), trailing-slash stripping, and the negative case
(model.base_url MUST NOT leak to non-active providers).
All 51 tests in the existing model-resolver + custom-provider banks still
pass.
Caught by maintainer review on stage-337 (full pytest with the new network
isolation in place surfaced the regression that the fork-CI mock-server path
would have hidden).
Tests should not reach the public internet. Before this commit, an
accidentally-leaking outbound socket from the test_server fixture (real
TLS handshakes to Anthropic / Amazon / OpenRouter, sometimes triggered
by SDK-init paths that found a credential the credential-strip allowlist
missed) was adding 60+s of wall-time to a 100s test run and creating a
class of flaky failures.
This installs a default-deny socket-block at two layers:
1. Pytest process, via tests/conftest.py module-level monkey-patch on
socket.create_connection + socket.socket.connect. Loopback / RFC1918
private / link-local / RFC2606 reserved-TLD destinations pass through;
anything else raises OSError("hermes test network isolation: outbound
to ... blocked"). Tests that legitimately need real outbound opt back
in via the new `allow_outbound_network` fixture (no current callers).
2. Test_server subprocess (server.py), via a HERMES_WEBUI_TEST_NETWORK_BLOCK=1
environment-variable-gated guard at the top of server.py. tests/conftest.py
sets the env var on every test_server spawn. Without this, the subprocess
could make outbound that the pytest-side block can't see (which is exactly
what was happening — verified via `ss -tnp` showing the server.py child
with established ESTAB sockets to [2607:6bc0::10]:443).
In production the env var is unset, so the guard is a no-op.
Companion changes:
- test_dns_resolution_failure refactored to mock socket.getaddrinfo
raising gaierror, instead of relying on a real DNS lookup of a
*.invalid hostname. The test was the one outlier that genuinely
exercised real DNS; mocking matches what every other probe-error test
in the same file already does.
- New tests/test_conftest_network_isolation.py with 9 adversarial
tests proving the block fires for public IPs (including the exact
Anthropic IPv6 and Amazon IPv4 destinations we observed leaking),
the allow-list passes loopback / RFC1918 / link-local / reserved-TLDs,
and the opt-in fixture re-enables real outbound when needed.
Test suite: 5,120 → 5,192 (+72 net new from this commit + the regression
tests in the companion commits). Wall time: 161s → 95s on the same
hardware. No remaining outbound from any test path.
Two test-infrastructure fixes surfaced while running the full suite on
this branch. Both prevent accidental outbound network calls from the
pytest process — a class of bug that doesn't show up as test failures
but corrupts timing, leaks credentials, and was responsible for a recent
10× slowdown observation.
## 1. AWS_EC2_METADATA_DISABLED for the whole pytest session
When hermes-agent's bedrock_adapter / botocore credential chain is
imported during tests (e.g. via api/config.py provider-catalog imports),
botocore probes the EC2 Instance Metadata Service at 169.254.169.254
looking for an instance role. On VPS hosts where IMDS is reachable but
rate-limited (HTTP 429) or non-responsive, those probes dominate wall
time — a 161s test run was observed extending to 600+s.
Set `AWS_EC2_METADATA_DISABLED=true` at module load (before any test-file
imports trigger botocore initialisation). This is the documented AWS-
supported way to silence the probe and matches the guard the agent's own
`hermes_cli/doctor.py` already uses inside its parallel-probe block.
Also explicitly re-set the var on the spawned test-server env so it
can't be accidentally cleared by a later `env.update(...)`.
## 2. Expanded credential-strip allowlist
The original strip list covered 6 providers (OpenRouter, OpenAI,
Anthropic, Google, DeepSeek, Xiaomi). Several others leaked through
into the test server subprocess:
- `MEM0_API_KEY`, `XAI_API_KEY`, `MISTRAL_API_KEY`, `OLLAMA_API_KEY`,
`GROQ_API_KEY`, `TOGETHER_API_KEY`, …
- AWS credentials (`AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`,
`AWS_SESSION_TOKEN`, `AWS_PROFILE`, `AWS_BEARER_TOKEN_BEDROCK`)
- Messaging bot tokens (`TELEGRAM_BOT_TOKEN`, `DISCORD_BOT_TOKEN`,
`SLACK_BOT_TOKEN`, `SIGNAL_API_TOKEN`, `WHATSAPP_API_TOKEN`)
- Memory providers (`HONCHO_API_KEY`, `SUPERMEMORY_API_KEY`)
- Search / browser / image-gen (`FIRECRAWL_API_KEY`, `FAL_KEY`,
`TAVILY_API_KEY`, `SERPER_API_KEY`, `BRAVE_API_KEY`)
- GitHub tokens (`GH_TOKEN`, `GITHUB_TOKEN`)
- Azure OpenAI (`AZURE_OPENAI_API_KEY`, `AZURE_OPENAI_ENDPOINT`)
A real outbound TLS connection to a provider's IPv6 endpoint was
observed during a test run on this host before the strip was expanded.
The test server uses a mock config and has no business making real API
calls.
## Test status
5,151 passed / 11 skipped / 1 xfailed / 2 xpassed / 0 regressions in
139s on Python 3.11. Down from 147s before the fixes (and from
intermittent 10×-slowdowns on IMDS-rate-limited hosts). All API/feature
contracts unchanged.
## Security audit of remaining test-suite host references
Every IP / URL / hostname referenced in `tests/**.py` was classified:
- Loopback (127.0.0.1, localhost, ::1, 0.0.0.0)
- RFC1918 private (10.*, 172.16-31.*, 192.168.*)
- RFC 5737 TEST-NET-3 documentation (203.0.113.*)
- RFC 2606 reserved docs domains (*.example.com, *.example.local,
*.example.test)
- Security-attack input strings used only as parser/validator input
(evil.com, attacker, evil.example.com — never resolved or contacted)
- Real provider/CDN endpoints used only as `base_url` config strings
or CSP-allowlist assertions — never actually fetched
- 8.8.8.8 used only as a "non-loopback example" in `_is_local_from_handler()`
unit tests
No suspicious egress destinations.
Lets desktop users collapse the session-list sidebar to maximise the chat
area, without adding any visible UI affordance. Default appearance is
identical to master — only users who actively try to toggle (or know the
keyboard shortcut) ever see a difference.
## Behaviour (desktop only, ≥641px)
| State | Action | Result |
|------------------------------------|-----------------------|-----------------------------------------|
| Sidebar open, click active rail | Toggle | Sidebar collapses to width:0 |
| Sidebar open, click different rail | Normal switch | **Sidebar stays open** (no surprise) |
| Sidebar collapsed, click any rail | Expand + switch | Sidebar expands, then panel switches |
| Anywhere, Cmd/Ctrl+B | Toggle | Same as same-active-rail click |
| Mobile (<641px), any of the above | No-op | Mobile overlay behaviour unchanged |
Two discoverability paths, both opt-in. **No new visible buttons.** Users
who never click the active rail icon see zero UI change vs. master.
## Surface-minimal design
The behaviour is contained behind one extra arg on the rail/sidebar-nav
onclick: `switchPanel('chat',{fromRailClick:true})`. Without that flag the
function preserves master's behaviour exactly — every programmatic
`switchPanel(name)` callsite (commands, deeplinks, internal state changes)
is unaffected. The guard chain inside `switchPanel`:
opts.fromRailClick && _isDesktopWidth() && (
_isSidebarCollapsed() ? expandSidebar() :
prevPanel === nextPanel ? (toggleSidebar(true); return false))
is the ONLY new code path that can cause a collapse. Cross-panel clicks
fall through to the existing switch logic untouched.
## Polish from both source PRs
- **Click-active gesture** as the primary toggle (#1884 @jasonjcwu — the
genuine UX innovation; no extra button needed)
- **Cmd/Ctrl+B keyboard shortcut** (#1924 @spektro33; VS Code convention).
Guarded against firing when typing in INPUT / TEXTAREA / contenteditable
so the shortcut never steals from in-progress text editing.
- **Inline flash-prevention `<script>`** in `<head>` (#1924) sets
`data-sidebar-collapsed='1'` on `<html>` BEFORE the stylesheet loads,
so cold loads with a persisted-collapsed state paint correctly from
frame 0 with no flicker. Cleared by JS once the class system takes over.
- **Smooth slide animation** via `.24s cubic-bezier(.22,1,.36,1)`
(#1924, mirrors the existing workspace-panel collapse on the right)
- **`aria-expanded` mirrored** on the active rail button (#1884) so
screen readers announce open/collapsed transitions.
- **`body.resizing` transition-suppression** (#1884) keeps the drag-resize
cursor instant — no animation during a width-resize gesture.
- **bfcache `pageshow` re-sync** (#1884) — if another tab toggled the
sidebar while this page was frozen, bring it in line on restore.
## Drops vs. #1924
- No persistent rail "toggle sidebar" button (Nathan: keep the UI stealth)
- No close-X button in chat panel head (same reason)
- No i18n keys for the dropped buttons
## What did NOT change
- 22 rail/sidebar-nav `onclick` handlers gained the `{fromRailClick:true}`
arg — function-call shape, invisible to users
- 1 inline `<script>` in `<head>` (flash prevention) — invisible
- 5 lines of CSS — invisible unless someone collapses
That's the entire visible-UI delta. **23 ins / 22 del on `index.html`,
all string-replace.**
## Verification
- 5,151 pytest passing including a new 34-test structural suite covering
every contract (CSS rules, JS functions, fromRailClick guard, legacy
proxy forwarding, flash-prevention `<script>` ordering, mobile
exclusion via :not(.mobile-open) selector, aria-expanded sync).
- Live browser walkthrough at 1280px verified:
- Default boot state identical to master (sidebar open, width 300px)
- Click active rail → collapse (width 1, opacity 0, translateX -14px,
localStorage='1', aria-expanded=false). Panel unchanged.
- Click active rail again → expand back to width 300, aria=true
- Click DIFFERENT rail → normal switch, sidebar stays open (legacy-
preserving case, verified explicitly)
- Click rail while collapsed → expand + switch in one gesture
- Cmd+B toggles correctly
- Cmd+B inside `<textarea>` → suppressed (defaultPrevented=false)
- Reload with collapsed state persisted → restores without flash
- Mobile simulation (matchMedia returns false for min-width:641px):
same-active-rail click is no-op, Cmd+B is no-op, sidebar stays at 300px
Co-authored-by: jasonjcwu <jasonjcwu@users.noreply.github.com>
Co-authored-by: spektro33 <spektro33@users.noreply.github.com>
Closes#1884Closes#1924
Two follow-ups from Opus pre-release review of stage-336:
1. tests/conftest.py — autouse session fixture that removes
HERMES_WEBUI_SKIP_ONBOARDING from os.environ for the whole pytest run, and
restores it after. Hosting providers and isolated harnesses set this var
to short-circuit the onboarding wizard, but it leaked into pytest and
caused tests that exercise apply_onboarding_setup() to fail with cryptic
FileNotFoundError. Tests that specifically validate the short-circuit
behavior can opt back in with monkeypatch.setenv. Surgical per-test
delenv calls remain as defense-in-depth but are now redundant.
2. docs/rfcs/README.md — one-line note that first-time contributor RFCs
should be discussed in an issue before opening a PR. Gates drive-by
design-doc PRs without us having to decline them on contribution.
Verified: 96 onboarding-related tests pass with HERMES_WEBUI_SKIP_ONBOARDING=1
exported in the test runner env (would have failed before this fixture).
1. test_issue1362_codex_oauth_onboarding.py::test_anthropic_onboarding_setup_allows_linked_oauth_without_api_key
Pre-existing env-collision bug, surfaced when HERMES_WEBUI_SKIP_ONBOARDING=1
is in the test runner env (set by hosting providers and by isolated test
harnesses). `apply_onboarding_setup()` short-circuits without writing the
config file when SKIP_ONBOARDING is set, but the test asserts the file was
written, so it fails with FileNotFoundError on read_text().
Fix: `monkeypatch.delenv("HERMES_WEBUI_SKIP_ONBOARDING", raising=False)` —
matches the convention already used in test_issue1499_keyless_onboarding.py
and test_issue1500_lmstudio_env_var_alignment.py.
2. test_issue1800_file_html_interactions.py::test_media_html_inline_keeps_csp_sandbox
Slicing-based source-string assertion (4000-char window after `def _handle_media`)
broke because PR #2044's MEDIA_ALLOWED_ROOTS parsing was inserted earlier in
the function and pushed the CSP block to offset 4211. Widened window to 5000.
Assertion content is structural (CSP sandbox string present), not positional.
Moves docs/turn-journal-rfc.md → docs/rfcs/turn-journal.md, establishing
the convention for future design documents on hermes-webui's data-at-rest
and recovery surfaces. Adds docs/rfcs/README.md describing when an RFC
applies (large changes, durability/recovery semantics, new infrastructure
primitives) and the simple status header convention.
Polish on turn-journal.md:
- Added 3-line status header (Status / Author / Created) at top.
- Light tone edits on two flourishes that read fine in a PR description
but felt off in permanent repo documentation. Author's voice preserved
throughout the rest of the document.
Co-authored-by: ai-ag2026 <261867348+ai-ag2026@users.noreply.github.com>
Two concrete data-corruption vectors flagged in Opus review of PR #2041,
both fixed atomically so the new repair-safe endpoint is safe for production:
1. Shared tmp filename under concurrent calls
`tmp = target.with_suffix('.json.reconcile.tmp')` produced a fixed path
per session ID. Two simultaneous repair-safe POSTs would interleave bytes
in the same tmp file, then both rename → corrupted JSON. Now matches the
`Session.save()` convention at api/models.py:484 with a pid+tid suffix.
2. TOCTOU between target.exists() check and tmp.replace(target)
`os.replace()` overwrites unconditionally. If a concurrent Session.save()
for the same SID materialized the live sidecar in the microsecond window
between the existence check and the rename, the reconciliation would
silently overwrite a live sidecar with a (lossier) state.db reconstruction.
Switched to `os.link()` + `unlink(tmp)` which is atomic create-or-fail —
on FileExistsError we record `skipped: sidecar_appeared_during_reconcile`
and keep the live sidecar untouched.
Plus a round-trip schema-parity test: materialize a sidecar from state.db,
then load it back through `Session.load()` and assert the messages survive.
Catches future schema drift between `_state_db_row_to_sidecar()` and
`Session.__init__()`. Also adds a guard test confirming the .reconcile.tmp
suffix includes pid+tid (regression guard for hazard #1).
Tests: 23 passing across the recovery suite (was 21; +2 new in this commit).
Co-authored-by: ai-ag2026 <261867348+ai-ag2026@users.noreply.github.com>
nesquena-hermes