| Files | Last commit | Last update |
|---|---|---|
refactor(devlog): publish devlog as tracked files and retire the private submodule devlog was a private submodule (lidge-jun/opencodex-internal) pinned by gitlink. The bookkeeping outgrew the benefit: 1723 commits touched the pointer, and dev, preview, and main each held a different one, so every branch move and promotion dragged a diff. The privacy was doing exactly one job — keeping pre-disclosure security material out of public view — and that job is now done by a rule plus two mechanical checks. WHAT MOVED OUT FIRST Two units of OPEN security review were excised to scratch before anything else was touched: _plan/260730_open_pr_backlog_triage and _plan/260730_new_issue_pr_triage, 12 tracked files. Both carried unresolved review verdicts on UNMERGED work, so no public diff exists to read them from. AGENTS.md already forbade that material in devlog/ — these were policy violations before publication was ever considered. The two documents an earlier pass had flagged instead (280 multi-auth patch plan, 145 common-security-hardening) turned out to be safe: all six fix commits named in the 280 manifest are ancestors of public main, with subjects that name the weakness class outright. Their NO-GO language described June, not now. THE CONVERSION - .gitmodules deleted; devlog was its only stanza - .gitignore devlog stanza replaced with the seven path rules ported from devlog's own ignore file, which stopped applying once devlog joined this repository. Without them a future `git add -A` would pull 129MB of vendored LiteLLM/CCA source. - gitlink removed, 1616 files added as ordinary blobs, 0 gitlinks left in the index - devlog/.git was moved to scratch rather than deleted, so the private-remote binding is recoverable PRIVACY SCAN NOW COVERS DEVLOG Removing devlog/ from EXCLUDED_PREFIXES is what makes "public" also mean "scanned". It surfaced 1254 findings, resolved by narrowing allowances rather than re-excluding: - home paths: the maintainer's own account name only, under devlog/ only. Any other username still fails anywhere — a contributor's home path is somebody else's data. - token-looking: shape-based, a token must SAY it is fake. A real high-entropy key matches none of those words. - emails: git-attribution CONTEXT rather than an address list. A Co-authored-by trailer, an `author Name <addr>` citation, or a table row carrying a 7+ hex SHA is provenance already public as commit authorship. A bare address pasted as contact detail still fails, and a new contributor needs no scanner change. Verified not-dead: a probe carrying a live-shaped key, a foreign home path, and a third-party email was staged and all three were caught. TWO CHECKS REPLACE THE PRIVACY tests/repo-hygiene.test.ts now asserts no gitlink anywhere, devlog tracked as blobs, no .gitmodules, vendored clones untracked, and excised triage untracked. Both new guards were driven RED before being trusted. A tripwire flags any OPEN _plan document combining an unresolved review verdict with security-boundary vocabulary. It took two corrections to become real: first it flagged the documents that DEFINE it, then — with that exemption in place — it PASSED while a genuinely offending document was staged, because the vocabulary list was English-only and this devlog is written in mixed Korean and English. Extended to both languages and re-driven red. Any future content check here must assume a bilingual corpus. Gates: typecheck clean, privacy:scan passed with devlog in scope, repo-hygiene 10/10. Full documentation of the decision trail, including the corrections this work made against its own earlier conclusions, is in devlog/_plan/260730_devlog_publication_feasibility/ (000-004, 010, 020, 030). In-progress GUI and src work in the working tree was deliberately left unstaged. | 3 months ago | |
perf(responses): reduce terminal repair and credential masking work (#8774) * perf(responses): reuse incremental buffer for terminal repair * docs(responses): record bounded comparison for PR 8774 * perf(redaction): defer ASCII offsets until a pattern matches * perf(redaction): reject impossible credential searches before slicing * docs(responses): record profiled masking fix and base comparison * docs(responses): expand preload path in comparison commands * test(responses): strengthen credential prefix coverage * test(redaction): mask minted credentials on the unframed ASCII path The literal-ASCII fast path in redactSecretString skips label folding. Minted OpenCodex credentials (#8794) carry no label, so their mask must run on both paths; this pins that for admin, data-plane and Remote Workspace tokens. --------- Co-authored-by: JUN <jun@lidgeai.com> | 2 days ago | |
feat(images,videos,codex): Grok Imagine 2.0/Video 1.5 bridges and quota/reset parity (#9013) * docs(devlog): plan Grok media and quota parity against live xAI/OpenAI behavior * feat(images): default the Grok image bridge to grok-imagine-image-2.0 and send quality grok-imagine-image-quality was retired on 2026-11-02 and is served as 2.0 at low quality, so the old default ignored the requested quality. The bridge and the /v1/images relay now default to grok-imagine-image-2.0, forward quality (low/standard -> low, medium/high/hd -> medium, auto) to 2.0 only, accept the full xAI aspect-ratio set, map 1024x1536/1536x1024 to 2:3/3:2, forward up to five edit sources, and return xAI validation errors only as canonical messages rebuilt from a closed vocabulary. * feat(videos): support Grok Imagine Video 1.5, 1080p and pinned frames The video bridge defaulted to the classic grok-imagine-video (720p, no frame pins). It now defaults to grok-imagine-video-1.5-lite, advertises only what the configured model accepts (1.5: 1080p, first/last frame, four keyframes, 14 references capped at 720p; Lite: first frame and 1080p; classic: first frame and seven references; unknown models keep the old surface), forwards image, last_frame, keyframes and reference_images in xAI's shape, and reports xAI validation errors and failure codes only as canonical messages. Image inputs are https or data:image URLs, or references to images the image bridge downloaded in this process, which resolve to the recorded xAI URL; no local file is read, and those references are only sent to the pinned xAI host. * fix(images): bound quality error details and align edit-source limits Review follow-up for the image bridge. A quality-list validation message is returned only with at most the four known values, each once, rebuilt in canonical order. The /v1/images/edits relay drops duplicate sources before the five-source limit, matching callXaiImages, and the guides now say the limit applies to that relay while the image_gen bridge tool sends one image. * fix(videos): refuse unsupported video settings and keep safe poll detail Review follow-up for the video bridge. An explicit resolution or aspect ratio the configured model cannot honour is now a tool error instead of being dropped, which would have bought a video at the upstream default. A permanent poll failure keeps the allowlisted xAI detail (attached as a separate field by the client) and still never the raw upstream message. * fix(codex): keep non-expiring reset credits and calibrate Pro 100/500 capacity The official Codex client types a reset credit's expires_at as optional, but the management DTO dropped any credit without a string expiry. Such rows are now kept with expires_at: null, together with the display fields the backend sends (reset_type, status, title, description, bounded) and total_earned_count; ids stay private. The auto-redeemer, which schedules by expiry, still sees only expiring credits. Pool capacity weighted Pro 100 (prolite) at the pre-2026-10-29 5x and had no Pro 500 (promax), which counted as one Plus seat. They are now 10x and 50x Plus, per OpenAI's 2026-10-29 doubling and staff-published multipliers. * docs(images): record remaining Grok/OpenAI non-LLM decisions and fix the video default comment * fix(gui): show non-expiring Codex reset credits as "No expiry" The reset-credit endpoint now returns expires_at: null for credits without an expiry, which the dashboard passed to the date formatter as the Unix epoch, so such a credit read "Expires Jan 1, 1970 (0d left)" and was marked urgent. The credit item now shows a localized "No expiry" line without urgency; dated credits are unchanged. * fix: guard reset-credit auto-redeem order, refuse invalid video durations Final-review follow-up. Consume names no credit and the dashboard says the oldest is used first. Now that non-expiring credits are visible, the auto-redeemer also receives them and stands down ("order-not-guaranteed") when a credit granted before the one it protects could be spent in its place. An explicit video duration outside 1-15 whole seconds (or a non-numeric one) is a tool error instead of being clamped or dropped before a paid call; the numeric string form xAI accepts ("8") still works. The pool-quota regression expectations follow the Pro 100 weight of 10 (weighted weekly 71 % instead of 66 %). * fix(codex): auto-redeem fails closed on tied or unorderable reset-credit grants An equal grant time has no known consume tie-breaker and a missing or unparseable one cannot be ordered, so either now stops automatic redemption like an older credit does. The redeemer's inspect reads every upstream row, passing a row without a grant time on as unorderable; the display DTO is unchanged. * fix(codex): never plan auto-redeem for a reset credit without a usable grant time A credit whose grant time is missing or unparseable has no identity the redeemer can protect, so it is never selected; it still blocks other plans as an unorderable competitor. The guard also rejects an unparseable planned grant time before checking competitors. * fix: treat identical reset credits as a tie, scope quality to 2.0, advertise all image ratios Final-approval review follow-up. - Auto-redeem excludes exactly one row as the planned credit; a second row with the same grant and expiry is a distinct credit with an unknown consume order, so the redeemer stands down. - quality is forwarded to grok-imagine-image-2.0 only; another 2.x model has to opt in explicitly rather than receive a field it may reject on a paid call. - The image_gen tool advertises every aspect ratio the xAI client forwards (adds 9:19.5, 19.5:9, 9:20, 20:9, 1:2, 2:1). * fix(images): refuse an edit relay request with any unusable image source Final-approval review follow-up: a source the relay cannot represent as a URL (an empty string, a file_id entry, an object without a URL, a non-string) used to be dropped while the remaining sources went to a paid edit. The whole request is now a 400 before any upstream call. * fix(images): validate every edit relay source as an https or image data URL Final-approval review follow-up: an explicit null source was still ignored and any non-empty string was forwarded. Each supplied source (null included) must now be an https URL without credentials or a base64 PNG/JPEG/WebP data URL, otherwise the request is a 400 before any upstream call. * fix(images): check every URL property of an edit entry and require canonical base64 Final-approval review follow-up: an images[] entry contributed only its first URL-bearing property, so {image_url: null, url} slipped through and a second valid property was dropped; and any run of base64 characters (====) passed as an image data URL. Each image_url / image_url.url / url property is now a source in its own right, and a data URL needs whole base64 quads with padding only at the end. * fix(images): require canonical base64 padding bits in edit data URLs Final-approval review follow-up: a payload such as AB== passed the alphabet, length and padding-position checks although its unused bits are non-zero. The payload must now round-trip through decode and re-encode unchanged. * fix(videos): refuse mode-specific settings xAI cannot honour (CodeRabbit) CodeRabbit follow-up. Before any paid call the video tool now also refuses: 1080p with last_frame or keyframes (they make it reference-to-video, which stops at 720p, like references already did); a keyframe outside the effective clip when duration is omitted (xAI's 8 s default); image plus references on the classic model, which rejects the mix; and an aspect ratio on plain image-to-video, where xAI keeps the image's own ratio. An xAI error body that stalls or is aborted while its detail is read no longer hides the HTTP status: detail extraction is best effort, so a permanent poll failure (e.g. 401) still stops polling instead of being retried as transient. * docs(videos): translate the video_gen parameter and model tables in all locales CodeRabbit follow-up: the eight translated video-bridge guides carried the new parameter rows, the refusal note and the model-table headers in English. * fix(videos): reject colliding keyframe slots and over-long classic reference clips (CodeRabbit) xAI snaps keyframes to a 1/3-second grid and rejects two in one slot, and the classic model limits reference-to-video to 10 seconds. Both are now refused before the paid submit; 1.5 references keep 15 seconds. | 23 hours ago | |
refactor(devlog): publish devlog as tracked files and retire the private submodule devlog was a private submodule (lidge-jun/opencodex-internal) pinned by gitlink. The bookkeeping outgrew the benefit: 1723 commits touched the pointer, and dev, preview, and main each held a different one, so every branch move and promotion dragged a diff. The privacy was doing exactly one job — keeping pre-disclosure security material out of public view — and that job is now done by a rule plus two mechanical checks. WHAT MOVED OUT FIRST Two units of OPEN security review were excised to scratch before anything else was touched: _plan/260730_open_pr_backlog_triage and _plan/260730_new_issue_pr_triage, 12 tracked files. Both carried unresolved review verdicts on UNMERGED work, so no public diff exists to read them from. AGENTS.md already forbade that material in devlog/ — these were policy violations before publication was ever considered. The two documents an earlier pass had flagged instead (280 multi-auth patch plan, 145 common-security-hardening) turned out to be safe: all six fix commits named in the 280 manifest are ancestors of public main, with subjects that name the weakness class outright. Their NO-GO language described June, not now. THE CONVERSION - .gitmodules deleted; devlog was its only stanza - .gitignore devlog stanza replaced with the seven path rules ported from devlog's own ignore file, which stopped applying once devlog joined this repository. Without them a future `git add -A` would pull 129MB of vendored LiteLLM/CCA source. - gitlink removed, 1616 files added as ordinary blobs, 0 gitlinks left in the index - devlog/.git was moved to scratch rather than deleted, so the private-remote binding is recoverable PRIVACY SCAN NOW COVERS DEVLOG Removing devlog/ from EXCLUDED_PREFIXES is what makes "public" also mean "scanned". It surfaced 1254 findings, resolved by narrowing allowances rather than re-excluding: - home paths: the maintainer's own account name only, under devlog/ only. Any other username still fails anywhere — a contributor's home path is somebody else's data. - token-looking: shape-based, a token must SAY it is fake. A real high-entropy key matches none of those words. - emails: git-attribution CONTEXT rather than an address list. A Co-authored-by trailer, an `author Name <addr>` citation, or a table row carrying a 7+ hex SHA is provenance already public as commit authorship. A bare address pasted as contact detail still fails, and a new contributor needs no scanner change. Verified not-dead: a probe carrying a live-shaped key, a foreign home path, and a third-party email was staged and all three were caught. TWO CHECKS REPLACE THE PRIVACY tests/repo-hygiene.test.ts now asserts no gitlink anywhere, devlog tracked as blobs, no .gitmodules, vendored clones untracked, and excised triage untracked. Both new guards were driven RED before being trusted. A tripwire flags any OPEN _plan document combining an unresolved review verdict with security-boundary vocabulary. It took two corrections to become real: first it flagged the documents that DEFINE it, then — with that exemption in place — it PASSED while a genuinely offending document was staged, because the vocabulary list was English-only and this devlog is written in mixed Korean and English. Extended to both languages and re-driven red. Any future content check here must assume a bilingual corpus. Gates: typecheck clean, privacy:scan passed with devlog in scope, repo-hygiene 10/10. Full documentation of the decision trail, including the corrections this work made against its own earlier conclusions, is in devlog/_plan/260730_devlog_publication_feasibility/ (000-004, 010, 020, 030). In-progress GUI and src work in the working tree was deliberately left unstaged. | 3 months ago | |
refactor(devlog): publish devlog as tracked files and retire the private submodule devlog was a private submodule (lidge-jun/opencodex-internal) pinned by gitlink. The bookkeeping outgrew the benefit: 1723 commits touched the pointer, and dev, preview, and main each held a different one, so every branch move and promotion dragged a diff. The privacy was doing exactly one job — keeping pre-disclosure security material out of public view — and that job is now done by a rule plus two mechanical checks. WHAT MOVED OUT FIRST Two units of OPEN security review were excised to scratch before anything else was touched: _plan/260730_open_pr_backlog_triage and _plan/260730_new_issue_pr_triage, 12 tracked files. Both carried unresolved review verdicts on UNMERGED work, so no public diff exists to read them from. AGENTS.md already forbade that material in devlog/ — these were policy violations before publication was ever considered. The two documents an earlier pass had flagged instead (280 multi-auth patch plan, 145 common-security-hardening) turned out to be safe: all six fix commits named in the 280 manifest are ancestors of public main, with subjects that name the weakness class outright. Their NO-GO language described June, not now. THE CONVERSION - .gitmodules deleted; devlog was its only stanza - .gitignore devlog stanza replaced with the seven path rules ported from devlog's own ignore file, which stopped applying once devlog joined this repository. Without them a future `git add -A` would pull 129MB of vendored LiteLLM/CCA source. - gitlink removed, 1616 files added as ordinary blobs, 0 gitlinks left in the index - devlog/.git was moved to scratch rather than deleted, so the private-remote binding is recoverable PRIVACY SCAN NOW COVERS DEVLOG Removing devlog/ from EXCLUDED_PREFIXES is what makes "public" also mean "scanned". It surfaced 1254 findings, resolved by narrowing allowances rather than re-excluding: - home paths: the maintainer's own account name only, under devlog/ only. Any other username still fails anywhere — a contributor's home path is somebody else's data. - token-looking: shape-based, a token must SAY it is fake. A real high-entropy key matches none of those words. - emails: git-attribution CONTEXT rather than an address list. A Co-authored-by trailer, an `author Name <addr>` citation, or a table row carrying a 7+ hex SHA is provenance already public as commit authorship. A bare address pasted as contact detail still fails, and a new contributor needs no scanner change. Verified not-dead: a probe carrying a live-shaped key, a foreign home path, and a third-party email was staged and all three were caught. TWO CHECKS REPLACE THE PRIVACY tests/repo-hygiene.test.ts now asserts no gitlink anywhere, devlog tracked as blobs, no .gitmodules, vendored clones untracked, and excised triage untracked. Both new guards were driven RED before being trusted. A tripwire flags any OPEN _plan document combining an unresolved review verdict with security-boundary vocabulary. It took two corrections to become real: first it flagged the documents that DEFINE it, then — with that exemption in place — it PASSED while a genuinely offending document was staged, because the vocabulary list was English-only and this devlog is written in mixed Korean and English. Extended to both languages and re-driven red. Any future content check here must assume a bilingual corpus. Gates: typecheck clean, privacy:scan passed with devlog in scope, repo-hygiene 10/10. Full documentation of the decision trail, including the corrections this work made against its own earlier conclusions, is in devlog/_plan/260730_devlog_publication_feasibility/ (000-004, 010, 020, 030). In-progress GUI and src work in the working tree was deliberately left unstaged. | 3 months ago | |
refactor(devlog): publish devlog as tracked files and retire the private submodule devlog was a private submodule (lidge-jun/opencodex-internal) pinned by gitlink. The bookkeeping outgrew the benefit: 1723 commits touched the pointer, and dev, preview, and main each held a different one, so every branch move and promotion dragged a diff. The privacy was doing exactly one job — keeping pre-disclosure security material out of public view — and that job is now done by a rule plus two mechanical checks. WHAT MOVED OUT FIRST Two units of OPEN security review were excised to scratch before anything else was touched: _plan/260730_open_pr_backlog_triage and _plan/260730_new_issue_pr_triage, 12 tracked files. Both carried unresolved review verdicts on UNMERGED work, so no public diff exists to read them from. AGENTS.md already forbade that material in devlog/ — these were policy violations before publication was ever considered. The two documents an earlier pass had flagged instead (280 multi-auth patch plan, 145 common-security-hardening) turned out to be safe: all six fix commits named in the 280 manifest are ancestors of public main, with subjects that name the weakness class outright. Their NO-GO language described June, not now. THE CONVERSION - .gitmodules deleted; devlog was its only stanza - .gitignore devlog stanza replaced with the seven path rules ported from devlog's own ignore file, which stopped applying once devlog joined this repository. Without them a future `git add -A` would pull 129MB of vendored LiteLLM/CCA source. - gitlink removed, 1616 files added as ordinary blobs, 0 gitlinks left in the index - devlog/.git was moved to scratch rather than deleted, so the private-remote binding is recoverable PRIVACY SCAN NOW COVERS DEVLOG Removing devlog/ from EXCLUDED_PREFIXES is what makes "public" also mean "scanned". It surfaced 1254 findings, resolved by narrowing allowances rather than re-excluding: - home paths: the maintainer's own account name only, under devlog/ only. Any other username still fails anywhere — a contributor's home path is somebody else's data. - token-looking: shape-based, a token must SAY it is fake. A real high-entropy key matches none of those words. - emails: git-attribution CONTEXT rather than an address list. A Co-authored-by trailer, an `author Name <addr>` citation, or a table row carrying a 7+ hex SHA is provenance already public as commit authorship. A bare address pasted as contact detail still fails, and a new contributor needs no scanner change. Verified not-dead: a probe carrying a live-shaped key, a foreign home path, and a third-party email was staged and all three were caught. TWO CHECKS REPLACE THE PRIVACY tests/repo-hygiene.test.ts now asserts no gitlink anywhere, devlog tracked as blobs, no .gitmodules, vendored clones untracked, and excised triage untracked. Both new guards were driven RED before being trusted. A tripwire flags any OPEN _plan document combining an unresolved review verdict with security-boundary vocabulary. It took two corrections to become real: first it flagged the documents that DEFINE it, then — with that exemption in place — it PASSED while a genuinely offending document was staged, because the vocabulary list was English-only and this devlog is written in mixed Korean and English. Extended to both languages and re-driven red. Any future content check here must assume a bilingual corpus. Gates: typecheck clean, privacy:scan passed with devlog in scope, repo-hygiene 10/10. Full documentation of the decision trail, including the corrections this work made against its own earlier conclusions, is in devlog/_plan/260730_devlog_publication_feasibility/ (000-004, 010, 020, 030). In-progress GUI and src work in the working tree was deliberately left unstaged. | 3 months ago |
devlog
Planning and investigation notes for lidge-jun/opencodex, tracked in that repository. Public.
Why these are public
These notes were a private submodule until the bookkeeping outgrew the benefit:
1723 commits touched the gitlink, and dev, preview, and main each held a
different pointer, so every branch move and promotion dragged a diff. The
privacy was doing one job — keeping pre-disclosure security material out of
public view — and that job is now done by a rule plus a check instead, which is
cheaper and does not tax every ordinary commit.
What that trade requires is stated below and is not optional.
What belongs here
Closed records: plans, audits, measurements, and post-mortems for work that has
shipped. A _fin/ unit documents something already visible in public git
history, so publishing the writeup discloses nothing the diff does not.
Layout
_plan/— units still open. One directory per unit, decade-numbered docs._fin/— closed units. A unit moves here once its terminal outcome is recorded (DONE,NOOP,BLOCKED, orNEEDS_HUMANwith the reason)._chase/— external reference material captured for parity comparisons.
Security working notes do not live here
Never commit unreleased security findings, draft advisories, exploit paths, or pre-disclosure patch plans here. This directory is public, so a commit is a disclosure the moment it is pushed, and the history is impractical to purge. A private repository is no better: it is cloned across machines and CI and outlives the embargo.
This binds maintainers. It has been violated by maintainer-authored triage
before — two units of open security review accumulated under _plan/ and had to
be excised before this directory could be published.
The deciding question before writing a security note here: is there already a
public diff that reveals this weakness? If yes, it is a closed record and
belongs in _fin/. If no, it is pre-disclosure material and belongs in scratch.
Security work happens in an untracked scratch directory (.tmp/ in the public
working tree, or a mktemp -d path), and only the published outcome — release
notes, advisory text, the fix itself — reaches any repository. See the
"Security working notes" section of the public AGENTS.md
for the full rule.
Submodule hygiene
The public repository pins this repository by commit and declares
ignore = dirty, update = none, shallow = true. Public CI never checks it
out and nothing in the build or test path reads from it, so an unavailable or
stale pointer cannot fail a public check. Do not add embedded git repositories
here; a nested .git becomes a 160000 gitlink in the parent index and breaks
actions/checkout for everyone.