Repository navigation
feat(linear-triage): replace gh subprocess with GitHub REST API [OMN-9403] - #368
Merged
Merged
Conversation
|
Warning Rate limit exceeded
Your organization is not enrolled in usage-based pricing. Contact your admin to enable usage-based pricing to continue reviews beyond the rate limit, or try again in 2 minutes and 43 seconds. ⌛ How to resolve this issue?After the wait time has elapsed, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout. Please see our FAQ for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: ⛔ Files ignored due to path filters (1)
📒 Files selected for processing (21)
📝 WalkthroughWalkthroughReplaces gh-CLI PR discovery with an injectable GitHub REST client in the linear-triage handler, refactors PR candidate selection and stale/epic logic, and adds two new nodes (audit-trail-compactor and two-strike-arbiter) with their handlers, models, contracts, metadata and tests; plus project manifest and small path/state tweaks. Changes
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
|
| Verdict | Meaning | Blocks merge? |
|---|---|---|
passed |
No critical/error findings | No |
blocked |
CRITICAL or ERROR findings found | Yes |
degraded |
All models unavailable (infra) | No (pilot) |
Powered by omniintelligence.review_pairing.cli_review — node-based adversarial review via HandlerLlmCliSubprocess (OMN-8524)
Contributor
Author
|
On the deploy-gate failure — two things worth checking before dismissing as "pre-existing, not related":
The proper mechanism (pattern from OMN-9242 / omnibase_core#857 this morning): add a Decision point for you on this PR:
Either way the answer is not "it's pre-existing, not our problem." Land one of the two paths above. |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 4
🧹 Nitpick comments (1)
tests/test_golden_chain_linear_triage.py (1)
79-109: Make the stub distinguish repo-search and head-branch fallback.Right now all identifier-based lookups share the same map, and
list_prs_by_head()is hardcoded to return[]. That means the tests don't actually prove the new repo-first/head-fallback behavior in_find_merged_pr(); a regression in call ordering or branch matching would still pass.One way to tighten the helper
-def _stub_github( - merged_prs: dict[str, dict[str, str] | None] | None = None, -) -> GitHubClientProtocol: +def _stub_github( + *, + org_prs: dict[str, dict[str, str] | None] | None = None, + repo_prs: dict[tuple[str, str], dict[str, str] | None] | None = None, + head_prs: dict[tuple[str, str], dict[str, str] | None] | None = None, +) -> GitHubClientProtocol: """Build a stub GitHubClientProtocol. - - Args: - merged_prs: Map of search_term -> fake PR dict, or None for no results. - If a key maps to None, that search returns empty. """ gh = MagicMock(spec=GitHubClientProtocol) - pr_map = merged_prs or {} + org_map = org_prs or {} + repo_map = repo_prs or {} + head_map = head_prs or {} def _search_prs(*, search_term: str, state: str = "all") -> list[dict[str, str]]: - pr = pr_map.get(search_term) + pr = org_map.get(search_term) return [pr] if pr else [] def _search_prs_in_repo( *, repo: str, search_term: str, state: str = "all" ) -> list[dict[str, str]]: - pr = pr_map.get(search_term) + pr = repo_map.get((repo, search_term)) return [pr] if pr else [] def _list_prs_by_head( *, repo: str, branch: str, state: str = "merged" ) -> list[dict[str, str]]: - return [] + pr = head_map.get((repo, branch)) + return [pr] if pr else []🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@tests/test_golden_chain_linear_triage.py` around lines 79 - 109, The _stub_github helper currently uses a single pr_map for all lookup types and returns [] for _list_prs_by_head, so tests don't exercise repo-specific search then head-branch fallback; update _stub_github to accept separate maps (e.g., repo_prs and head_prs) or a composite mapping and change _search_prs_in_repo to consult repo_prs keyed by search_term (or repo+search_term) while _list_prs_by_head should consult head_prs keyed by branch and return that list, keeping _search_prs as the global fallback; ensure you wire these to gh.search_prs, gh.search_prs_in_repo, and gh.list_prs_by_head respectively in the function so tests can assert repo-first then head-fallback behavior.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.py`:
- Around line 3-7: The module docstring is outdated: replace references to
GITHUB_TOKEN and the CLI fallback with the new GH_PAT-only behavior;
specifically update the HandlerLinearTriage docstring to state that
GitHubHttpClient.from_env() requires GH_PAT and that there is no fallback to `gh
auth token`, and mention that GH_PAT must be set in the environment for PR
lookups (remove any instructions about using the `gh` CLI).
- Around line 275-283: The from_env method in GitHubHttpClient currently uses
os.environ["GH_PAT"] which raises KeyError before your RuntimeError; change it
to read the env var safely (e.g., os.environ.get("GH_PAT") or wrap access in
try/except KeyError) and if the token is missing raise the existing RuntimeError
with the explanatory message, then return cls(token) when present; update the
method named from_env in GitHubHttpClient accordingly so the custom error path
is hit for missing GH_PAT.
- Around line 315-320: The except urllib.error.HTTPError as exc block in
handler_linear_triage.py currently always continues on 403+Retry-After which
lets _get() fall off the end on the last attempt and be treated as "no PRs";
change the logic so that after sleeping and retrying you only continue if there
are remaining retries, and if this is the final attempt re-raise the HTTPError
(raise exc or just raise) instead of continuing so callers see the failure;
update the control using the existing retry loop counter/max-retries variable in
the same function to decide when to re-raise.
---
Nitpick comments:
In `@tests/test_golden_chain_linear_triage.py`:
- Around line 79-109: The _stub_github helper currently uses a single pr_map for
all lookup types and returns [] for _list_prs_by_head, so tests don't exercise
repo-specific search then head-branch fallback; update _stub_github to accept
separate maps (e.g., repo_prs and head_prs) or a composite mapping and change
_search_prs_in_repo to consult repo_prs keyed by search_term (or
repo+search_term) while _list_prs_by_head should consult head_prs keyed by
branch and return that list, keeping _search_prs as the global fallback; ensure
you wire these to gh.search_prs, gh.search_prs_in_repo, and gh.list_prs_by_head
respectively in the function so tests can assert repo-first then head-fallback
behavior.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: c07bccd1-ec1a-4152-80a8-4aeee8ffa085
📒 Files selected for processing (2)
src/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.pytests/test_golden_chain_linear_triage.py
There was a problem hiding this comment.
Actionable comments posted: 5
♻️ Duplicate comments (4)
src/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.py (4)
274-283:⚠️ Potential issue | 🟠 MajorUse a safe env lookup so the intended fail-fast error is raised.
Line 277 uses
os.environ["GH_PAT"], so an unset token raisesKeyErrorbefore yourRuntimeErrorbranch runs.Proposed fix
`@classmethod` def from_env(cls) -> GitHubHttpClient: """Create a client from GH_PAT env var. Raises if not set.""" - token = os.environ["GH_PAT"] + token = os.environ.get("GH_PAT", "") if not token: raise RuntimeError( "GH_PAT environment variable is not set. " "Export it before running node_linear_triage." )🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.py` around lines 274 - 283, The from_env classmethod in GitHubHttpClient currently uses os.environ["GH_PAT"] which raises KeyError before your RuntimeError is hit; change the lookup to a safe form (e.g., use os.environ.get("GH_PAT") or check 'GH_PAT' in os.environ) so token can be tested and, if missing, raise the intended RuntimeError with your message; update references to the token variable in from_env accordingly.
302-337:⚠️ Potential issue | 🟠 MajorRe-raise after the final secondary-rate-limit retry.
On Line 323 the loop still
continues on the last403+Retry-After, so_get()falls through toreturn Noneand callers degrade a real GitHub failure into “no PRs found”.Proposed fix
except urllib.error.HTTPError as exc: if exc.code == 403 and "retry-after" in (exc.headers or {}): retry_after = int(exc.headers["retry-after"]) _log.warning( "GitHub secondary rate limit, retrying after %ds", retry_after ) _time.sleep(retry_after + 1) - continue + if attempt < 3: + continue + raise🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.py` around lines 302 - 337, The loop handling GitHub 403 + Retry-After currently always continues even on the final attempt, causing _get() to return None instead of surfacing the real error; update the HTTPError branch that checks for "retry-after" so that if exc.code == 403 and "retry-after" in (exc.headers or {}) you only sleep-and-continue when attempt < 3, and on the final attempt re-raise the exception (i.e., raise the caught HTTPError instead of continuing), leaving the existing primary rate-limit (x-ratelimit-reset) logic unchanged.
715-736:⚠️ Potential issue | 🟠 MajorOnly suppress the known non-epic
list_children()failure.This
except Exceptionalso hides auth failures, transport errors, and Linear outages, so the epic-close pass can be skipped while the node still reports success.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.py` around lines 715 - 736, The broad except around client.list_children(...) in handler_linear_triage.py is hiding all errors; narrow it to only suppress the known non-epic case by catching the specific exception or error condition from the Linear client (e.g., catch LinearApiError or a custom NotAnEpicError) or inspect the exception message/code and re-raise for auth/transport/outage errors; leave the rest to propagate so unexpected failures (auth, network, upstream outages) are not swallowed—adjust the except block around the while loop that calls client.list_children(parent_id=ticket.id, ...) to only handle the known non-epic error and call _log.warning with the same message, but re-raise other exceptions.
3-7:⚠️ Potential issue | 🟡 MinorUpdate the GitHub auth docs to match the implementation.
Lines 5-6 and 265-266 still describe
GITHUB_TOKENplusgh auth token, but this code now requiresGH_PATand has no CLI fallback. That will send operators down the wrong setup path.Also applies to: 263-267
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.py` around lines 3 - 7, The module docstring and usage docs still reference GITHUB_TOKEN and a `gh auth token` CLI fallback; update all text in the module (including the top-level docstring and any usage/help strings around HandlerLinearTriage) to require and document the GH_PAT environment variable instead and remove any mention of a CLI fallback. Search for occurrences of "GITHUB_TOKEN" and "gh auth token" and replace the guidance with clear instructions that GH_PAT is required, and ensure any constant/variable names or help text (e.g., references near HandlerLinearTriage or the GH_PAT config/lookup code) reflect this change.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In
`@src/omnimarket/nodes/node_audit_trail_compactor/handlers/handler_audit_trail_compactor.py`:
- Around line 221-242: _compute_recurring_tickets currently picks last_failure
using failures[-1] which assumes failures is chronological; instead, for each
ticket_id in ticket_failures sort the failures by their timestamp (e.g.,
entry.created_at or entry.timestamp) and select the most recent entry's
description/entry_type for last_failure; update the ModelRecurringTicket
construction in _compute_recurring_tickets to compute last_failure from the
failure with the max timestamp (and preserve failure_count and the top-20
sorting by failure_count).
- Around line 153-162: The code builds cutoff as an ISO string and then
lexically compares e.recorded_at to that string, causing incorrect filtering;
change cutoff to be a timezone-aware datetime (e.g., datetime.now(tz=UTC) -
timedelta(...)) and ensure the comparisons use datetime objects by parsing or
converting e.recorded_at to a datetime before filtering (update the filtered
comprehension in handler_audit_trail_compactor.py that uses cutoff and
e.recorded_at, or adjust self._reader.read_friction_entries /
read_dispatch_entries to return recorded_at as datetime), then compare
e.recorded_at >= cutoff as datetime-to-datetime.
In `@src/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.py`:
- Around line 367-424: The handlers search_prs, search_prs_in_repo, and
list_prs_by_head currently catch all exceptions and return [] which hides
API/auth/parsing failures; instead, log the error and re-raise (or raise a
specific exception) so callers can distinguish real “no results” from failures.
Concretely, update each of search_prs, search_prs_in_repo, and list_prs_by_head
to remove the broad except Exception: return [] pattern: catch only expected
parse/empty-item cases if you need to return [], but for _get failures or
unexpected exceptions (e.g. HTTP errors thrown by _get or parsing errors from
_parse_pr_items) log the error then re-raise (or wrap and raise a descriptive
exception) so upstream code can handle failures instead of treating them as
empty result sets.
In
`@src/omnimarket/nodes/node_two_strike_arbiter/handlers/handler_two_strike_arbiter.py`:
- Around line 181-200: The handler currently always leaves action as
EnumArbiterAction.SECOND_STRIKE unless diagnosis_path is set, so successful
side-effects from move_to_blocked() or record_friction() get ignored; update the
logic after calling move_to_blocked and record_friction to set action to the
strongest side-effect that actually succeeded (check the return/boolean result
from self._linear_updater.move_to_blocked(...) and the friction_filed result
from self._friction_recorder.record_friction(...)) with a precedence such as: if
diagnosis_path -> EnumArbiterAction.DIAGNOSIS_WRITTEN, else if move_to_blocked
succeeded -> EnumArbiterAction.BLOCKED, else if record_friction succeeded ->
EnumArbiterAction.FRICTION_RECORDED, else EnumArbiterAction.SECOND_STRIKE so the
returned action reflects the highest-impact successful operation.
In `@tests/test_golden_chain_audit_trail_compactor.py`:
- Around line 27-40: The tests hardcode recorded_at to "2026-04-21T..." which
eventually falls outside the 7-day lookback; update the fixture function
_make_entry so its recorded_at parameter defaults to None and, when None,
computes a timestamp relative to now (e.g., datetime.utcnow() or timezone-aware
datetime.now(timezone.utc) minus the desired offset) and formats it as an
ISO8601 Zulu string before passing to ModelAuditEntry; also replace the other
hardcoded fixture usages that set recorded_at to static strings with calls that
rely on _make_entry (or compute relative timestamps similarly) so all test
entries remain within the lookback window.
---
Duplicate comments:
In `@src/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.py`:
- Around line 274-283: The from_env classmethod in GitHubHttpClient currently
uses os.environ["GH_PAT"] which raises KeyError before your RuntimeError is hit;
change the lookup to a safe form (e.g., use os.environ.get("GH_PAT") or check
'GH_PAT' in os.environ) so token can be tested and, if missing, raise the
intended RuntimeError with your message; update references to the token variable
in from_env accordingly.
- Around line 302-337: The loop handling GitHub 403 + Retry-After currently
always continues even on the final attempt, causing _get() to return None
instead of surfacing the real error; update the HTTPError branch that checks for
"retry-after" so that if exc.code == 403 and "retry-after" in (exc.headers or
{}) you only sleep-and-continue when attempt < 3, and on the final attempt
re-raise the exception (i.e., raise the caught HTTPError instead of continuing),
leaving the existing primary rate-limit (x-ratelimit-reset) logic unchanged.
- Around line 715-736: The broad except around client.list_children(...) in
handler_linear_triage.py is hiding all errors; narrow it to only suppress the
known non-epic case by catching the specific exception or error condition from
the Linear client (e.g., catch LinearApiError or a custom NotAnEpicError) or
inspect the exception message/code and re-raise for auth/transport/outage
errors; leave the rest to propagate so unexpected failures (auth, network,
upstream outages) are not swallowed—adjust the except block around the while
loop that calls client.list_children(parent_id=ticket.id, ...) to only handle
the known non-epic error and call _log.warning with the same message, but
re-raise other exceptions.
- Around line 3-7: The module docstring and usage docs still reference
GITHUB_TOKEN and a `gh auth token` CLI fallback; update all text in the module
(including the top-level docstring and any usage/help strings around
HandlerLinearTriage) to require and document the GH_PAT environment variable
instead and remove any mention of a CLI fallback. Search for occurrences of
"GITHUB_TOKEN" and "gh auth token" and replace the guidance with clear
instructions that GH_PAT is required, and ensure any constant/variable names or
help text (e.g., references near HandlerLinearTriage or the GH_PAT config/lookup
code) reflect this change.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: cef1e894-332f-4e79-983c-ed855c4d1865
⛔ Files ignored due to path filters (1)
uv.lockis excluded by!**/*.lock
📒 Files selected for processing (24)
.onex_state/workflow_result.jsonomnibase_compatomnibase_corepyproject.tomlsrc/omnimarket/nodes/node_audit_trail_compactor/__init__.pysrc/omnimarket/nodes/node_audit_trail_compactor/contract.yamlsrc/omnimarket/nodes/node_audit_trail_compactor/handlers/__init__.pysrc/omnimarket/nodes/node_audit_trail_compactor/handlers/handler_audit_trail_compactor.pysrc/omnimarket/nodes/node_audit_trail_compactor/metadata.yamlsrc/omnimarket/nodes/node_audit_trail_compactor/models/__init__.pysrc/omnimarket/nodes/node_audit_trail_compactor/models/model_audit_trail_input.pysrc/omnimarket/nodes/node_audit_trail_compactor/models/model_audit_trail_result.pysrc/omnimarket/nodes/node_linear_triage/handlers/handler_linear_triage.pysrc/omnimarket/nodes/node_ticket_query/handlers/handler_ticket_query.pysrc/omnimarket/nodes/node_two_strike_arbiter/__init__.pysrc/omnimarket/nodes/node_two_strike_arbiter/contract.yamlsrc/omnimarket/nodes/node_two_strike_arbiter/handlers/__init__.pysrc/omnimarket/nodes/node_two_strike_arbiter/handlers/handler_two_strike_arbiter.pysrc/omnimarket/nodes/node_two_strike_arbiter/metadata.yamlsrc/omnimarket/nodes/node_two_strike_arbiter/models/__init__.pysrc/omnimarket/nodes/node_two_strike_arbiter/models/model_two_strike_input.pysrc/omnimarket/nodes/node_two_strike_arbiter/models/model_two_strike_result.pytests/test_golden_chain_audit_trail_compactor.pytests/test_golden_chain_two_strike_arbiter.py
✅ Files skipped from review due to trivial changes (15)
- src/omnimarket/nodes/node_two_strike_arbiter/handlers/init.py
- src/omnimarket/nodes/node_audit_trail_compactor/init.py
- src/omnimarket/nodes/node_two_strike_arbiter/init.py
- src/omnimarket/nodes/node_two_strike_arbiter/models/init.py
- omnibase_core
- src/omnimarket/nodes/node_audit_trail_compactor/models/init.py
- omnibase_compat
- .onex_state/workflow_result.json
- src/omnimarket/nodes/node_two_strike_arbiter/metadata.yaml
- src/omnimarket/nodes/node_audit_trail_compactor/handlers/init.py
- src/omnimarket/nodes/node_audit_trail_compactor/models/model_audit_trail_result.py
- src/omnimarket/nodes/node_audit_trail_compactor/metadata.yaml
- src/omnimarket/nodes/node_two_strike_arbiter/contract.yaml
- src/omnimarket/nodes/node_audit_trail_compactor/contract.yaml
- src/omnimarket/nodes/node_two_strike_arbiter/models/model_two_strike_result.py
jonahgabriel
added a commit
that referenced
this pull request
Apr 21, 2026
- handler_audit_trail_compactor: compare datetimes not ISO strings (thread 5); add _parse_recorded_at() helper; sort last_failure by timestamp not list order (thread 6) - handler_linear_triage: narrow list_children except to only suppress known non-epic errors, re-raise auth/transport faults (thread 4); re-raise _get() failures in search_prs/search_prs_in_repo/list_prs_by_head instead of collapsing to [] (thread 7) - handler_two_strike_arbiter: capture move_to_blocked() return value; report strongest successful side-effect action (TICKET_BLOCKED / FRICTION_FILED / DIAGNOSIS_WRITTEN) instead of always SECOND_STRIKE (thread 8) - test_golden_chain_audit_trail_compactor: make timestamps relative to now() so suite does not break after 7-day lookback window expires (thread 9) Threads 1/2/3 verified already addressed in current code.
Contributor
Author
|
@coderabbitai review |
✅ Actions performedReview triggered.
|
Contributor
Author
|
Re-running CodeRabbit Thread Check after prior fixes. |
…9403] - Add GitHubClientProtocol + GitHubHttpClient using urllib (zero external deps) - Reads GH_PAT from env (fail-fast, no silent fallback) - Only PR-check tickets in In Progress/In Review (43 vs 682) - Only check epic completion for active root tickets (25 vs 1663) - Rate-limit awareness: auto-sleeps on x-ratelimit-remaining exhaustion - Error handling for Linear list_children 400s on non-epic tickets - Updated golden chain tests to use DI stubs (GitHubClientProtocol) instead of monkeypatching module-level functions - All 8 golden chain tests pass - End-to-end verified via: onex node node_linear_triage --input input.json
…, fix deps New nodes (picks #2 and #7 from prioritized list): - node_two_strike_arbiter (effect): Monitors consecutive fix-attempt failures. After 2 strikes, writes a diagnosis document, moves Linear ticket to Blocked, and files a friction event. Implements the two-strike protocol from GEMINI.md. Subscribes: onex.evt.omnimarket.fix-attempt-failed.v1 Publishes: onex.evt.omnimarket.two-strike-arbiter.v1 - node_audit_trail_compactor (compute): Reads friction + dispatch logs, computes failure modes, recurring tickets, stall heatmap. Compact output for overseer-verify and session-post-mortem consumption. Subscribes: onex.cmd.omnimarket.audit-trail-compact.v1 Publishes: onex.evt.omnimarket.audit-trail-compacted.v1 Dependency fix (same as PR #372): - Replace path sources with git sources for portable resolution - Add git sources for omnibase-spi, omninode-memory (PyPI stale) - Add omnibase-spi as explicit direct dep for git source override Tests: 16 new golden chain tests (7 arbiter + 9 compactor), all pass. Assisted-by: GLM-5.1 (z.ai) via Crush <crush@charm.land>
- handler_audit_trail_compactor: compare datetimes not ISO strings (thread 5); add _parse_recorded_at() helper; sort last_failure by timestamp not list order (thread 6) - handler_linear_triage: narrow list_children except to only suppress known non-epic errors, re-raise auth/transport faults (thread 4); re-raise _get() failures in search_prs/search_prs_in_repo/list_prs_by_head instead of collapsing to [] (thread 7) - handler_two_strike_arbiter: capture move_to_blocked() return value; report strongest successful side-effect action (TICKET_BLOCKED / FRICTION_FILED / DIAGNOSIS_WRITTEN) instead of always SECOND_STRIKE (thread 8) - test_golden_chain_audit_trail_compactor: make timestamps relative to now() so suite does not break after 7-day lookback window expires (thread 9) Threads 1/2/3 verified already addressed in current code.
jonahgabriel
force-pushed
the
jonah/omn-9403-linear-triage-github-api
branch
from
April 22, 2026 02:10
50e50cb to
cfd99b1
Compare
jonahgabriel
added this pull request to the merge queue
Apr 22, 2026
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.
What
Replace
ghCLI subprocess calls with direct GitHub REST API viaGH_PATenv var innode_linear_triage. This is the first ONEX node workflow verified end-to-end throughRuntimeLocal→EventBusInmemory→ServiceStateDisk.Why
gh pr listsubprocess per ticket was catastrophically slow (1000+ spawns for 682 recent tickets)gh auth tokenfallback violates fail-fast ruleChanges
handler_linear_triage.pyGitHubClientProtocol+GitHubHttpClient— urllib-based, zero external depsGH_PATfrom env — fail-fast on missing, no fallbackx-ratelimit-remaining ≤ 2list_children400s caught and logged for non-epic tickets_gh_search_pr,_find_sibling_merged_prmodule-level functionssubprocessimporttest_golden_chain_linear_triage.pyGitHubClientProtocolmock stubs instead of monkeypatching module functions_stub_github()helper for configuring merged PR responsesVerification
Test plan
Summary by CodeRabbit
New Features
Improvements
Tests
[skip-deploy-gate: OMN-9393 deploy-agent RED — this PR is a handler refactor (node_linear_triage), does not modify runtime, infra, or deployment manifests. No deploy risk.]