fix: replace Record<string, any> with Record<string, unknown> in GraphQL hooks - #1288
Open
Mayank Joshi (Lothnic) wants to merge 1 commit into
Open
fix: replace Record<string, any> with Record<string, unknown> in GraphQL hooks#1288Mayank Joshi (Lothnic) wants to merge 1 commit into
Mayank Joshi (Lothnic) wants to merge 1 commit into
Conversation
…hQL hooks - Remove stale TODOs from mediaUrls.ts (URLs already present) - Rename ambiguous type params T/V to TResult/TVariables - Replace Record<string, any> | undefined with Record<string, unknown> = Record<string, never> - Remove no-longer-needed biome-ignore suppressions
Contributor
|
PR author is not in the allowed authors list. |
Contributor
|
All contributors have signed the CLA. Thank you! |
Mayank Joshi (Lothnic)
changed the title
Author
|
I have read the CLA Document and I hereby sign the CLA |
Vasilev Dmitrii (gHashTag)
added a commit
to gHashTag/BrowserOS
that referenced
this pull request
Aug 22, 2026
…pe (browseros-ai#1287) The pull-request poller caught its error and dropped it. The error is now a required argument of the only emitter - it cannot be dropped again by construction - and the message names the shape: an HTTP status when the error carries one, the transport error otherwise, a decode failure as itself. Proven by running the drill on the branch before acceptance: HTTP 502, transport URLError -1009, and a decode failure that says the fetch discarded its status. That last line is the bee finding the next defect and naming it honestly; it is now browseros-ai#1288. Closes browseros-ai#1287
Vasilev Dmitrii (gHashTag)
added a commit
to gHashTag/BrowserOS
that referenced
this pull request
Aug 22, 2026
…could see The heuristic reads the acceptance criteria, extracts the identifiers they name, and calls the work done if all of them are present in the boundary files. That is sound for an issue that asks for something new - the identifier does not exist until the work exists. It is exactly backwards for a fix: browseros-ai#1288's criteria name fetchPullRequest BECAUSE that function loses the HTTP status, so its presence is the defect, and the Queen read it as the cure. Measured 2026-08-23: eight of seventeen candidates were skipped this way, including two issues filed that same hour. The board looked finished while half of it had never been touched. Its own comment already warned that a behavioural task's FILES exist from day one; the identifiers have the same problem one level down. Presence is evidence only if it postdates the issue. The issue's created_at now travels from the epic timeline through the store into the candidate, and one git log per boundary path asks whether anything has changed there since. Nothing changed means nothing done, whatever names the file contains. Asked only after the identifier check has already passed, so the common case never pays for it, and a path git cannot answer for counts as changed: a false "already done" loses the work entirely, while a false "not done" costs one turn in which the bee says it is done. Gate: REAL_EXIT=0. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Vasilev Dmitrii (gHashTag)
added a commit
to gHashTag/BrowserOS
that referenced
this pull request
Aug 22, 2026
…rowseros-ai#1288) fetchPullRequest read 'let (data, _)' and discarded the response, so a 404 for a deleted pull request and a 200 with an unreadable body reached the poller as the same silent decode error. The response is kept, the failure carries the status it arrived with, and nil is carried rather than turned into a zero a reader would take for an observation. The poller, the drill and a suite now all see it. Wire proof from the drill on this branch: 'HTTP 404: response did not decode as a pull request for /repos/gHashTag/trios/pulls/4242' beside its status-less twin, and queen.drill.poll_failure.passed. make e2e: 969 checks in 87 scenarios. Closes browseros-ai#1288
Vasilev Dmitrii (gHashTag)
added a commit
to gHashTag/BrowserOS
that referenced
this pull request
Aug 22, 2026
…rom the night the release lane stood still A delegated issue is a contract with three parts that must agree: the boundary, the criteria, and the proof. When they disagree the bee cannot win, and its failure reads as the bee's fault. I wrote two briefs on 2026-08-23 whose criteria demanded files the boundary forbade. browseros-ai#1287 asked for a logic suite and permitted only ChatViewModel, so the bee put its suite inside ChatViewModel - the only file it could write. browseros-ai#1288 asked for the poller's message, a drill and a suite, and permitted only the client: four of five criteria unreachable, and a bee whose own change was exact came back 1/5. Both had to be finished at review time, and the second issue says so in its closing comment, because a record that blames a bee for a boundary its author drew is a record that lies. The new skill carries that rule, the trap where criteria naming a defect's own identifiers make the issue read as already done, what 'proven by a run' requires of the boundary, a five-question checklist, and the three ways a short return can be read - only one of which is the bee's fault. Live forensics gains the night's four: a keychain that times out on every operation is a waiting dialog (pgrep -x SecurityAgent, and three discriminators that named it); 'open --env' applies only at launch, so kill and VERIFY dead first; a refusal you cannot find is usually under a key you did not grep - '/accept' logs queen.accept.blocked; and the shared trinity checkout breaks this build about once every two hours. Gate: REAL_EXIT=0. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
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.
Description:
CLA Signature
I have read the CLA Document and I hereby sign the CLA