fix(gateway): implement session-scoped /fast + raise KANBAN_GUIDANCE ceiling - #137
Closed
exiao wants to merge 2 commits into
Closed
fix(gateway): implement session-scoped /fast + raise KANBAN_GUIDANCE ceiling#137exiao wants to merge 2 commits into
exiao wants to merge 2 commits into
Conversation
…ceiling Two pre-existing baseline test failures on live-config, unrelated to any open feature PR (verified on a clean live-config worktree): 1. test_fast_command.py (3 tests): the /fast session-scope feature was half-shipped. The tests expect /fast to be session-scoped by default with a --global flag (mirroring /reasoning), using the already-present _session_service_tier_overrides dict and _resolve/_set helpers. But _handle_fast_command still only wrote agent.service_tier to config.yaml globally and rejected --global as unknown. Rewrite it to match the shipped /reasoning pattern: bare /fast records a session override and leaves config.yaml untouched; --global persists to config.yaml and clears the session override. 2. test_kanban_guidance_prompt_size_bounded: upstream b5bd0ef added legitimate attachment guidance, growing KANBAN_GUIDANCE to 5749 chars past the test's <5500 cap. The test's docstring says the ceiling guards unbounded growth 'with a little headroom', so restore headroom by raising the cap to 6500 (still a real guard). 122 tests pass across both files; ruff clean. Patch note: ~/.hermes/plans/hermes-patches/baseline-flaky-tests.md
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 0893678620
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Owner
Author
|
Superseded by #139, which consolidates the live-config CI baseline repair. #139 fixes the same failing tests via the correct root cause (deleting the stale shadow slash-handlers in gateway/run.py that were overriding the picker-enabled mixin versions) and, unlike the partial repairs, does not break test_send_file_tool. See #139 for the full diagnosis and the one remaining product decision (the egress-vs-send_file contradiction). |
exiao
added a commit
that referenced
this pull request
Jul 20, 2026
…e shadow handlers) (#139) * fix(gateway): delete stale shadow slash-handlers + raise KANBAN_GUIDANCE ceiling Consolidates the live-config CI baseline repair (supersedes #136, #137, #138). 15 tests were failing on live-config across 4 modules. This fixes 12 via the correct root cause; 3 egress tests are a separate product decision (see patch note). 1. Stale shadow handlers (11 tests: test_fast_command x3, test_choice_picker x8). The June slash-command extraction moved _handle_reasoning_command and _handle_fast_command into GatewaySlashCommandsMixin with the new choice-picker logic, but left the OLD copies in the GatewayRunner class body in run.py. A method defined on the subclass overrides an inherited mixin method, so the stale copies shadowed the picker-enabled versions, making the choice picker unreachable and /fast --global unrecognized. Delete the two stale defs (166 lines) so the mixin versions activate. #137's re-implement-into-the-shadow approach was wrong; the shadow must be deleted. 2. KANBAN_GUIDANCE size ceiling (1 test). Upstream b5bd0ef added legitimate attachment guidance (5749 chars vs the <5500 cap). Raise cap to 6500. 219 tests pass across the affected modules; send_file's 57 expanded-ext tests still pass (unlike #138 which broke them). ruff clean. Patch note: ~/.hermes/plans/hermes-patches/live-config-baseline-consolidated.md * fix(gateway): route MEDIA-tag code/config exts through validated egress Third baseline failure (test_platform_base.py::TestUniversalMediaEgress x3). The egress redesign (2fd36b1, NousResearch#36060) intends unknown/code extensions to deliver ONLY via the validated pass (validate_media_delivery_path: exists on disk, safe root, not denylisted) so a prompt-injection MEDIA:/etc/x.py stays visible instead of silently exfiltrating. But the code still folded MEDIA_TAG_EXTRA_EXTS (.py/.log/.toml/...) into _MEDIA_TAG_EXT_ALTERNATION, the UNCONDITIONAL extract grammar — contradicting its own design comment and the egress tests. Point _MEDIA_TAG_EXT_ALTERNATION at MEDIA_DELIVERY_EXTS only, so code/config/log tags route through the validated pass. MEDIA_TAG_EXTRA_EXTS is retained for the separate producer-tool detector in run.py (_TOOL_MEDIA_RE). test_send_file_tool's expanded-ext test used nonexistent /tmp paths, which under the validated design correctly no longer extract. Fixed the fixture to create a real file in a monkeypatched safe root, exercising the true intended delivery path for every extension. Verified: TestUniversalMediaEgress 13/13, test_send_file_tool 85/85, full affected set 426 passed. The 5 unrelated red tests (SSRF, url_safety, cross_profile_guard) are pre-existing on clean live-config and untouched here. Approach A per Eric: egress design authoritative. * fix(gateway): honor effective model for fast mode
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
1 participant
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.
Problem
origin/live-confighad 4 failing tests unrelated to any open feature PR. Confirmed pre-existing by running them on a cleanlive-configworktree with zero edits. They were dragging every open PR's CI red via the sharedPython testsslices.Fixes
1.
/fastsession-scope was half-shipped (3 tests)tests/gateway/test_fast_command.pywas updated to expect/fastto be session-scoped by default with a--globalflag (mirroring/reasoning), using the_session_service_tier_overridesdict and the_resolve_session_service_tier/_set_session_service_tier_overridehelpers — all of which already exist ingateway/run.py. But_handle_fast_commandwas never updated: it still only wroteagent.service_tierto config.yaml globally and rejected--globalas an unknown argument. The tests asserted behavior the handler didn't implement.Rewrote
_handle_fast_commandto match the shipped/reasoningpattern:--global(any position)._resolve_session_service_tier(session_key=...)./fast fast|normalrecords a session override and leaves config.yaml untouched./fast ... --globalwritesagent.service_tierto config.yaml AND clears the session override so the global default wins.No new env vars, no prompt-cache impact.
2.
KANBAN_GUIDANCEsize ceiling too tight (1 test)Upstream
b5bd0ef38("docs(kanban): port attachment guidance into KANBAN_GUIDANCE") added legitimate content, growing the block to 5749 chars against the test's< 5_500cap. The test's own docstring says the ceiling "guards against unbounded growth, not against any growth" and is sized "with a little headroom." Raised the cap to6_500to restore headroom (still a real unbounded-growth guard); lower bound unchanged.Verification
test_fast_command.py+test_kanban_tools.py: 122 passed.ruff checkclean on all changed files.Once merged, this un-reds the shared test slices for the other open PRs (#129, #131, #133).