* feat(cli): share rendered videos through unified reviews (#7135)
* feat(share): attach editable project sources to reviews (#7137)
* feat(share): measure local sharing outcomes (#7138)
* feat(cli): stage the sixty-day publish transition (#7136)
* feat(studio): share review pages from local preview (#7140)
Advertise the CLI sharing bridge before bootstrapping it so standalone Studio does not
request unsupported routes. Preserve origin and session checks.
Budget the intended sharing artifact: clean main packs at 8,060,519 bytes versus
8,162,979 with sharing (8,169,168 in CI). Required policy and Unicode tables, bundled
sharing code, and Studio controls account for the growth. Set the packed-only budget
to 8,577,627, following the 5% headroom convention; retain the unpacked cap and all
package checks.
* fix(share): persist refusals and improve CLI retry feedback
* fix(ci): restore render diagnostics and include sharing workspace
* fix(cli): keep sharing owner error codes stable
* fix(share): clarify update feedback and align audience changes
* fix(share): clarify updates and avoid scratch-file rerenders
* fix(cli): report active shares and preserve update channels
* fix(cli): resolve update channel when checking registry
* fix(ci): refresh sharing package budgets after rebase
* fix(share): require consent before uploading project files
* fix(parsers,lint,producer,cli): confine composition path lookups to the project
- The local asset resolver offers only files inside the project: a path
that is still absolute after its leading `../` is clamped (an encoded
`//`, a drive letter, a leading backslash) yields no candidate.
- Lint, the catalog-usage walk and `keyframes` read a
data-composition-src only inside the project. Lint and the catalog
walk resolve it with the render's own resolver, now shared from
`@hyperframes/parsers/asset-resolution`, and lint reports one outside
the project as the render pre-flight does.
- The render compile step does not look at a media, gsap script or image
path on another UNC or device root, and the GIF and HEIC preparation
treats a `//host/...` image src as a URL rather than a share.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(parsers,lint): keep Windows absolute asset paths that render copies in
The shared asset resolver now offers only files inside the project. On
Windows a drive-letter src such as C:\media\clip.mp4 used to reach the
real file through the old fallback, and render still copies that file in.
Lint now accepts such a file for media and for CSS written in the page,
so it raises no new missing_local_asset or texture_mask_asset_not_found
error. The media-treatment and normalize-audio commands find it again.
The check never looks at another computer's share, and on macOS and
Linux a src that starts with / is still a project URL, so nothing
changes there.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(parsers): expect a drive-letter sub-composition src to be refused on Windows
On Windows an absolute src outside the project is a drive-letter path,
which resolveCompositionSrc refuses, as the render pre-flight already
does. On POSIX the same src is a project URL and resolves inside the
project. The test now expects each platform's result.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(lint): do not report a missing timeline registry a script outside the project may hold
Lint no longer reads a script outside the project. When a root
composition loads such a script, lint cannot know whether the script
registers the timeline, so it no longer reports missing_timeline_registry
for that composition. Render still loads the script. A composition whose
scripts are all in the project is checked as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(capture): validate the Vertex region and model id before captioning
Vision captioning passes the region and model id to the Google SDK, which
builds the request host from the region and the request path from the model
id. A region that is not a plain Google Cloud region name, or a model id with
URL delimiters, now skips captioning with a warning instead of building a
request. Every real region (including "global") and model id still works.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(capture): accept colons inside Google model id segments
Vertex endpoint paths in domain-scoped projects contain a colon
(projects/example.com:proj/...). The colon cannot change the request host,
so allow it inside a segment. Leading slashes, empty segments, percent
signs, query and fragment delimiters stay rejected.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* test(cli): give each test file its own home so renders never wait on another file's lock
* test(cli): pin the cli test config after adding the per-file home setup
* feat(engine): decouple transparent background from png capture format
PNG capture used to imply a cleared page background. A new
CaptureOptions.transparentBackground lets lossless opaque PNG capture
(4:4:4 MP4 output) keep the page background. Absent, it still follows
format === "png", so alpha output and motion blur are unchanged.
Opaque PNG skips drawElement, whose verification and worker-encode
paths are JPEG-only.
Refs #5555
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* feat(engine): add 4:4:4 chroma option to getEncoderPreset
getEncoderPreset takes an optional chromaSubsampling ("420" by
default). "444" returns yuv444p for SDR MP4, so libx264 encodes the
High 4:4:4 Predictive profile and odd sizes need no even padding.
WebM, MOV and HDR presets ignore it. The ChromaSubsampling type is
exported for the producer.
Refs #5555
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* feat(producer): chromaSubsampling render option with lossless png capture
RenderConfig.chromaSubsampling ("420" by default) opts an SDR MP4
render into 4:4:4. "444" encodes libx264 yuv444p and captures every
frame as opaque PNG: Chrome's quality-100 JPEG is already 4:2:0 on
beginFrame and screenshot capture, and drawElement can fall back to
screenshot mid-render, so PNG is the only capture that keeps the
colour. drawElement is turned off for these renders.
The probe and main capture options share resolveCaptureSurface, so
they cannot disagree on format, quality or background.
Formats other than mp4, GPU encoding and HDR output are rejected
before any work starts, auto-detected HDR sources fail before capture,
and distributed requests refuse 444. The default render is unchanged.
Refs #5555
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* feat(producer,studio-server): accept chromaSubsampling on render routes
The shared parseEncoderOptions validates chromaSubsampling and answers
400 for "444" with a non-mp4 format, gpu, or force-hdr, so a request
that would come back 4:2:0 never starts a render. The Studio render
route, both Studio adapters and the producer HTTP server forward it.
The producer server also refuses "444" with outputDynamicRange "hdr",
which parseEncoderOptions cannot see.
No Studio UI control; the option is route-level like crf and hdrMode.
Refs #5555
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* feat(cli): add --chroma-subsampling flag
`hyperframes render --chroma-subsampling 444` opts an SDR MP4 render
into full-resolution chroma (yuv444p with lossless PNG capture). The
default stays 420. The plan rejects 444 with a non-mp4 format, --gpu
or --hdr before anything renders, with a usage error that names the
conflicting flag.
The value reaches the local render request and, only when it is 444,
the Docker argv, so default Docker invocations are unchanged.
Refs #5555
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* docs: document 4:4:4 chroma subsampling
Add a "4:4:4 masters" section to the rendering guide: when to use
--chroma-subsampling 444, why it captures PNG, the measured chroma gain
(large for text and UI, none for footage and gradients), its limits,
and a player compatibility table to be filled in from testing.
Document the flag in the CLI reference, the built-in rendering docs and
the hyperframes-cli skill reference (with the regenerated skills
manifest), and the chromaSubsampling option in the producer docs.
Refs #5555
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* docs: add 4:4:4 player compatibility results
Fill the 4:4:4 compatibility table from testing H.264 High 4:4:4
Predictive output. Chrome 155, Edge 155, VLC 3.0, Safari on iOS 26.6
and Safari 26.6 on macOS play it, and so does AVFoundation on macOS
26.6, the framework QuickTime Player plays through. Firefox 157 on
Windows and both Windows media players do not.
Refs #5555
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* docs: note the measured render cost of 4:4:4 output
Replace the vague "slower capture" note with measured numbers: a text
and UI composition rendered in about twice the default time, full-frame
footage in about three and a half times, because PNG encoding cost grows
with image detail.
Refs #5555
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(producer): reject chromaSubsampling 444 in distributed render config
Distributed configs have no chroma field and validateDistributedRenderConfig
ignored unknown keys, so a JS caller passing chromaSubsampling: "444" to the
Lambda or Cloud Run SDK got a 4:2:0 file with no error. Reject it, and make
the rendering guide say so precisely.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(producer): check a 444 render reaches the encoder as yuv444p
Every helper had tests, but nothing checked the orchestrator passes
chromaSubsampling through, so dropping it from the encoder preset or capture
surface would turn 444 renders back into 4:2:0 with all tests green. Render a
small fixture with "444" and with the default and check pix_fmt with ffprobe.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* refactor(producer): move the 444 drawElement gate into a helper
Keeps runCompileStage under the fallow complexity threshold.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore: keep comment share at or below the merge-base
comment-ratchet failed on 11 files this PR touched. Keep the 4:4:4
rationale once in chromaSubsampling.ts, move the distributed-render
reason into its error message, and cut the rest.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(producer): check a 444 render captures full-resolution chroma
The encoder check alone passes when chromaSubsampling is dropped from the
capture surface: frames are captured as JPEG and encoded yuv444p with
4:2:0 colour. Put a red/blue edge at an odd x, inside a 4:2:0 chroma
pair, and require the two pixels to differ in Cb.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore(producer): trim the chromaSubsampling.ts header to the package comment share
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: finish switching to capturesAlpha from #6262
The merge that adopted main's transparent / capturesAlpha left these files
uncommitted, so resolveCaptureSurface still returned transparentBackground.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* refactor(engine): use transparent / capturesAlpha from #6262
main added CaptureOptions.transparent and capturesAlpha (#6262), which do
what this PR's transparentBackground did. Use them, keep the PNG check in
drawElementWorkerEncodeEligible, and re-pin test-classification.mjs and
the skills manifest against current main.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore(producer): cut the 4:4:4 comment blocks to one to three lines
The reasoning lives in the PR description with the measurements.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* refactor: reconcile 4:4:4 chroma with latest main
Keep main's video-layer capture overrides, template entry resolution,
ProRes profile parsing and blur-differential tests alongside the chroma
option, and fold the encoder sub-parsers into one loop.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore: keep main's compile-stage tests and regenerate the skills manifest
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(studio-server): keep cache writes inside the project and canonicalize preview paths
- The tool's own folders (.hyperframes records, backups, preview documents,
.thumbnails, .transcode-cache) are no longer written, pruned or emptied
through a link that points outside the project. Thumbnails are still
served when their cache folder is a link; they are just not cached.
- The hf-id stamp record and the history id are replaced atomically instead
of written through an existing link.
- Sub-composition previews and thumbnails use the composition's canonical
project-relative path, and the path is attribute-encoded in the document.
- Thumbnail requests from another site are refused before any capture.
- Saved GSAP scripts escape a `</script` sequence so the script element
keeps its contents.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(core): refuse a UNC or device path on another root before any disk access
isSafePath now rejects a target whose root is a UNC or device root other than
the project's own before it calls realpath or lstat. Projects on a share keep working.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(studio): answer WebMCP clients only from Studio's own page
The polyfill now starts with its tab transport limited to Studio's origin and
its frame transport off, so a page that frames Studio cannot list or call its tools.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(inspector): check Host on the per-session preview server
The preview page server applies the inspector API's loopback Host rule, so a
rebound DNS name cannot read the bundled composition or project files.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* revert(studio-server): keep writing media proxies through a linked .transcode-cache
Refusing the folder made every proxied video in such a project answer 502.
A linked cache folder now receives copies again; it is still never swept,
so nothing already in the folder it points at is removed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore(studio-server): drop a comment the comment ratchet refuses
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): skip HEIF and GIF preview prep when .hyperframes/prepared-assets is a link
The preview transform wrote converted images into .hyperframes/prepared-assets
through a link, so a project shipping .hyperframes as a link to another folder
had files written there. Such a project now gets its preview with the images
as written in the composition, unconverted, and nothing is written.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(studio-server): never sweep a linked .transcode-cache, from Studio or hyperframes clean
The link check moves into cleanupProxyCache, which both Studio's cache
maintenance and `hyperframes clean` call, so neither removes files from the
folder a linked .transcode-cache points at.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(docs): escape catalog preview values and the install command's --vars argument
The catalog preview embeds variable values as JSON that cannot close a
script element or attribute, accepts value messages only from the page
that embeds it, and reads only declared ids with primitive values from
the link. The install command shell-quotes its --vars JSON and keeps
only variable-id keys with primitive values.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(core): fold CSS line endings before decoding escapes in the SVG sanitizer
A browser folds CR LF, CR and form feed into one newline before it reads
CSS escapes, so the whitespace after a hex escape can be a CR LF pair.
The sanitizer now does the same, so an escaped function name split by a
line ending is checked under its real name. The media-use bundle is
regenerated from the source.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(lint): match template ids and CSS rules in linear time
A getElementById template literal is matched against document ids with
an in-order indexOf scan instead of a regex of joined wildcards, so the
time no longer grows with the number of placeholders. The CSS rule
scanners only start a match at the beginning of a brace-free run, which
finds the same rules without rescanning long brace-less text. Lint
results on every HTML file in the repository are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(parsers): keep GSAP writer output inside its script element
String literals and quoted keys the GSAP writers emit now encode a
script closer or comment opener as \u003c, so a value can no longer end
the inline script that holds the timeline. Strings without those
sequences are written exactly as before. The writers also refuse a tween
method outside set, to, from and fromTo, and a motion path whose
coordinates or curviness are not numbers, since both are emitted as code.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(scripts): confine registry asset hosting to files inside the item
The asset hosting script now refuses a manifest path that leaves the
item directory or reaches its file through a symlink, before anything is
read, staged, uploaded or removed. A repository test fails when the
registry tracks a symlink or a manifest path that leaves its item.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(registry): pin the map blocks' d3 and atlas files to exact versions
The five map blocks load d3 7.9.2, the build the d3@7 range serves today,
with a subresource integrity hash, and fetch us-atlas 3.0.1 and
world-atlas 2.0.2, the files their ranges serve today. The bytes are
unchanged, so the blocks render the same. A repository test fails when a
registry item names an npm CDN package by range, tag or no version.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(registry): pin map blocks to d3 7.9.0
7.9.0 has been the stable release since 2024; 7.9.1 and 7.9.2 were published days ago. Pin the
long-standing version, with its subresource integrity hash.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(registry): keep the map block PR source-only and formatted
Restore the generated catalog pages and payloads, which the catalog
publish job regenerates from the item sources. Name the us-atlas URL
in us-map-flow so the call fits on one line and formatting is unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* refactor(scripts): check registry asset containment with one prefix test
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(lint): scan simple CSS rules in linear time
The two CSS-rule scans in the gsap rules took a body that could cross
'{', so a closed style of many '{' with no '}' rescanned to its end from
every brace. A scanner now returns the same selector and body pairs and
remembers the next '}', so each scan is linear. Findings are identical
on every tracked HTML file.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(core): label the form-feed sanitizer case as a guard
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): resolve helper programs from PATH, never the working directory
On Windows a bare command name is looked up in the working directory before
PATH. The CLI now sets NoDefaultCurrentDirectoryInExePath at startup, which
Node's spawn and cmd.exe both honor and child processes inherit, so tools
such as powershell.exe, taskkill and npx.cmd always come from PATH.
The shared ffmpeg/ffprobe resolver no longer scans the working directory on
Windows and skips relative PATH entries.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): keep Python helpers from importing files in the working directory
`python -c` and `python -m` put the working directory first on sys.path, so
a file in the project named like a probed module ran instead of the
installed package. Inline helper code now drops that entry before any
import, and `-m` helpers (pip) run through runpy after the same step. Paths
and arguments are unchanged.
On Windows, findPython also skips a `where` match in the working directory.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): run the skills installer and nested CLI calls independent of the project
The global skills install ran `npx skills@<pin>` in the current folder, so
npm could take the package or registry from that project's node_modules or
.npmrc. It now runs from the home folder; the install is global either way.
When the CLI starts the media-use engine it hands over its own entry file.
The local voice fallback and the audio helpers then run that same CLI with
the same Node instead of resolving `npx hyperframes` from the project.
Without a handed-over entry they behave as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(media-use): load only provider keys from a project .env
The audio scripts copied every key of a project .env into their environment
except two. They now take the same kind of short allowlist the CLI uses:
HeyGen, ElevenLabs and Gemini API keys plus a few harmless settings. Program
overrides, runtime options, credential files and the credential store path
come only from the person's own environment.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): build whisper.cpp only from a pinned, verified release
The source-build fallback cloned the upstream default branch. It now clones
the v1.9.5 release tag and refuses to build unless the checkout is the
reviewed commit for that tag.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): keep registry caches apart and refuse config install targets
Registry cache file names now include a digest of the exact registry URL,
and each entry records the URL it came from, so one registry's cached
manifest is never read for another. Entries from older versions are
refetched once.
Registry installs refuse targets inside .git or any hidden directory, hidden
configuration files, and package.json, hyperframes.json and
hyperframes.lock.json. An example installed by `init` may still bring its
own package.json and .gitignore.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(capture): keep page text to one line in agent files and validate Vertex keys
Capture writes AGENTS.md, CLAUDE.md and .cursorrules from the page's title,
font names and colours. Page text is now flattened to one short line and
colours must be hex values. Existing agent files that capture did not write
are kept, with a warning.
HYPERFRAMES_VERTEX_SERVICE_ACCOUNT must now be a plain service-account key
with a Google token endpoint; other credential types skip captioning with a
warning.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(cli): find the digest-named registry cache and trim a new file's comment
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): match registry targets as Windows names them; skills install without HOME
Windows drops trailing dots and spaces from a path segment and reads name:stream as
name, so package.json. or .git./hooks reached the files the target rules refuse.
The global skills install now falls back to the temp folder when HOME does not exist.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): turn off the Windows working-directory lookup in every program entry
The warm-check server and client are their own tsup entries and copied the
environment before cli.js set NoDefaultCurrentDirectoryInExePath, so the server
spawned a bare powershell.exe that Windows could take from the project folder.
Every entry that hides child consoles now also loads windowsExeSearch first.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): refuse Windows short-name registry targets
A segment like CLAUDE~1 or PACKAG~1.JSO opens the long name on NTFS, so it
reached the files the target rules refuse. The trailing-dot and stream-name
normalization is removed: publication already rejects those segments.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): lambda destroy runs SAM in a private folder and acts only on the selected stack
- sam delete outside a checkout runs in a fresh private temp folder,
removed afterwards, instead of the shared temp directory.
- A stack state file must name the stack it is selected by.
- destroy and deploy derive the video repository name from the stack
name instead of trusting the recorded value.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): snapshot keeps remote clips in a private folder and reads only allowed URLs
- A video URL the download policy refuses (plain http, or a private,
loopback or link-local host) is no longer handed to FFmpeg; the clip
is skipped instead, as render does.
- Remote clips are cached in a per-user, owner-only temp folder; a
folder open to other users is never reused.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): upgrade --project writes package.json through an exclusive temp file
The rewrite now uses the shared atomic replace, which creates a fresh,
unpredictable sibling with exclusive creation, so an existing
package.json.tmp entry is never written through.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): keep env files, Git data and tool records out of served and uploaded projects
- play and present answer 404 for .env files, .git and the tools'
private .hyperframes records, as Studio's preview does for the records.
- publish and cloud render never stage an env file or tool record, even
when markup references it, and never pull in an env file from outside
the project.
- Cloud site uploads (Lambda and Cloud Run deploySite) leave env files out.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): lambda commands accept a state file whose stack name differs only in case
On a case-insensitive disk, deploy --stack-name=Demo and --stack-name=demo share one
state file, so a name that matches apart from case is the same stack the user deployed
and keeps working as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): snapshot's fallback media folder uses createStagingDir
The repository check forbids bare mkdtemp beside a caller-supplied path, because it
fails past 260 characters on Windows. Also trims a comment so the new helper file stays
within its package's comment share.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): snapshot never hands FFmpeg a remote URL
FFmpeg follows redirects and resolves hosts on its own, so a clip the downloader could
not fetch must not reach it as a URL. Snapshot now leaves such a clip out and warns,
as render does. A fallback media folder is removed when the process exits.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): cover .envrc, .env-local and .env_prod, and test present's private-file 404
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(studio-server): thumbnail origin from the server, sandbox previews from other sites
The thumbnail route builds the headless browser's URL from the server's own
listening origin (a new optional `serverOrigin` on the adapter, wired in the CLI
preview server and the Vite dev host) instead of the request's Host header.
Hosts that report no origin fall back to a loopback Host only.
Preview responses (bundled, sub-composition and project files) opened from
another site carry `Content-Security-Policy: sandbox allow-scripts` and
`Vary: Sec-Fetch-Site`. Studio's same-origin iframe, browser automation and
non-browser clients are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(studio): build the thumbnail origin from a hostname bind name
With HYPERFRAMES_PREVIEW_HOST set to a hostname, the listening address is
the IP it resolved to, and the Host check accepts the bind name, not that
IP. The thumbnail browser then loaded a 403 page. Use the bind name when it
is a hostname; IP and wildcard binds keep the listening address.
— Rames Jusso
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* style(studio): drop the serverOrigin comment that put it over the package's comment share
The hostname-bind rule is stated by its test name in serverOrigin.test.ts.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(producer): a render uses the CLI's own GSAP, so the default CDN's file needs no network
* fix(cli): the GSAP vendor lookup runs as ESM from source, so render starts again
* chore(producer): raise the unpacked size budget by the bundled-script module's 3 KB
A project-local node_modules/.bin/hyperframes can shadow plain npx hyperframes,
but npm install -g never updates it. Classify the shadowing copy like the
running one: global installs keep their manager's command, project-local
installs get the project's own manager (from its lockfile), and unknown
layouts stay quiet. The command pins the release the npx copy is running.
* fix(reverse): give agents the sign-in reason to say, and catch stale CLIs that know reverse
- `reverse start`'s signed-out `next` now carries a sentence the agent can say as is
(why HeyGen, and that the HyperFrames account is the HeyGen account), asks for it at
the start of the sign-in message, and on remote terminals has the agent send the
device link and code before it waits.
- references/start.md: same ordering, an example that fits both login routes, and no
links to or quotes from the instructions themselves.
- An old global install that knows `reverse start` but predates `next` was never caught:
agent runs get no update notice and no auto-install, so it stays frozen. The skill now
also treats `_meta.updateAvailable: true` from `reverse start` as an old install and
runs the rest with `npx -y hyperframes@latest`; every CLI that knows `reverse start`
already reports that field.
* fix(reverse): send remote terminals straight to device sign-in, reason in every link message
- `reverse start` now knows when no browser can open (the check `auth login` uses, moved
to attendedTerminal): `login_command` is `auth login --device` and `next` asks for one
message with the reason, the link and the code. The two-step route (a refused browser
login, then a bare follow-up with the code) is where the reason got lost.
- Every message that gives the sign-in link starts with the reason; agents say they are
waiting rather than quoting the skill, and give no file paths.
- The out-of-date line for an old global install goes in the next message to the person.
* fix(reverse): flag an older global install that shadows npx, and pin the run's release
- installerDetection: `shadowingInstall` finds the older global `hyperframes` that a plain
`npx hyperframes` runs instead of this npx copy (first on PATH outside the npx cache), with
the command that moves it. `detectInstaller` takes the entry to classify.
- `reverse start` run through npx opens `next` with that line as words to say, signed in or
not. Skill prose alone did not get agents to tell the person.
- Skill: a signed-out `reverse start` with no `reverse.next` is also an old install, which
covers a cold update cache; the run pins the release `@latest` resolved to.
* fix(studio-server): reject cross-origin state-changing requests
Non-GET routes in createStudioApi answer 403 unless the shared
sameOriginPost predicate (lifted into studio-server's ./request-identity
subpath) accepts the request: the Host must be a server identity, a
present Sec-Fetch-Site must be same-origin, and a present Origin must
match the effective host and scheme. Hosts behind Codespaces or tunnels
allow their public name with HYPERFRAMES_PREVIEW_ALLOWED_HOSTS, read by
`hyperframes preview` and the Studio dev server.
Squashed from #6128 (42574a7, 110a300, 89011e1, 41c6626, 27193bb) onto
current main.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): the preview server answers only the hosts it is reached by
Every route of `hyperframes preview` (the forwarded studio API, the SSE
stream, /__hyperframes_config and the Studio page) now answers only
loopback names, the HYPERFRAMES_PREVIEW_HOST bind, or a host listed in
HYPERFRAMES_PREVIEW_ALLOWED_HOSTS, the same list the write guard uses.
Only the Host header counts here, since a page can set X-Forwarded-Host
on its own requests and a read carries no Origin to compare it with.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Dextheking1 <Dextheking1@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat(cli): label GPT image prices as estimates and show the charged credits
GPT Image generations on a HyperFrames plan are moving to fal's reported
vendor cost, so the pre-run figure is an estimate. The --plan basis for
gpt-image-2.5, 2.6 and 2.6-flare now reads "about N credits (charged at
vendor cost)". Nano Banana is unchanged.
When the /v3/images job carries charged.amount, generate reports it as
charged_credits and records it in the media ledger's provenance. Without
it, nothing new is printed. --max-credits is unchanged.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(cli): keep 10% headroom under --max-credits for vendor-priced images
gpt-image-2.5, 2.6 and 2.6-flare are charged at the vendor's cost, so
--max-credits now refuses one when its estimate plus 10%, rounded up to a
whole credit, is over the limit. A 16:9 gpt-image-2.5 image (about 40
credits) is refused at --max-credits 40 and passes at 44. Other models keep
the plain estimate check. The refusal and the --max-credits help say so, and
the help notes this is not a hard ceiling.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): count only HyperFrames-credit charges, and keep headroom under the run's limit
Review fixes for the vendor-cost image work:
- charged_credits is reported only when the job's charged.pool is
hyperframes_credits. A charge to api_credits or any other pool is in
that pool's own unit, so an API-key caller no longer sees it as HF
credits.
- runJob logs that charged amount, not the estimate, on a job's completion
row in heygen-jobs.jsonl, so later credit-limit checks see the real total.
Without a HyperFrames-credit charge the estimate is kept.
- The run's credit limit (.hyperframes/credit-limit.json) uses the same 10%
headroom rule as --max-credits for gpt-image-2.5, 2.6 and 2.6-flare,
through one shared limitCheck helper, with the same refusal wording.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* docs(media-use): say the credit limit is not a hard ceiling for GPT Image
A GPT Image call is admitted at its estimate plus 10% but charged the
vendor's cost, so it can still go past a run's credit limit. The media-use
skill, its HeyGen reference, the --max-credits help and the limit doc
strings now say so, and that a finished call's reported charge counts
toward the limit afterwards. Other models keep the "refuses what would go
past it" wording. No logic change.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): hold a running GPT Image call at its headroom under the project limit
A reserved or still-running vendor-priced call was logged at its plain
estimate, so two concurrent calls could each pass the 10% headroom check
and together go over. Log the headroom amount until HeyGen reports the
charge.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): templates pull unpacks under .hyperframes so a watched project can take the template
* fix(cli): templates pull's staging comment fits on one line
createSvgContactSheet built each path with path.join and then took the
label with split("/"), so on Windows the label was the whole path and
the 60-character cell limit cut the file name off. Use basename, as the
scroll and snapshot sheets in the same file already do.
* fix(cli): staging folders beside a project path no longer fail past 260 characters on Windows
mkdtemp does not get the long-path prefix that mkdir does, so every staging folder
made beside a project or output path failed once its full path passed 260 characters.
They now come from one helper, createStagingDir in @hyperframes/core/atomic-file, and a
scan test fails on a new mkdtemp beside a path.
* fix(cli): give mkdtemp the long-path form instead of replacing it
CodeQL treats mkdtemp as the safe way to make a folder under the system temp dir and
flagged every file written into a folder from the mkdir-based helper. createStagingDir
now calls mkdtemp with the long-path form of the path on Windows, so both hold.
* fix(cli): leave non-Windows paths alone when making a staging folder
The long-path form is now chosen by the shape of the path, not process.platform,
so tests that pretend to be Windows keep working and the comment share does not rise.
* fix(core): a staging folder keeps a Windows path that already names its namespace
* fix(core): a staging folder keeps a Windows device path as the caller gave it
* fix(core): the staging folder helper's comment fits the line limit
* fix(cli): no macOS developer-tools prompt from git, xcrun or cc checks on a Mac without the tools
* fix(cli): treat Xcode.app as installed tools and only take executables from PATH
* test(cli): give skills tests a developer folder so darwin cases run on any host
* fix(cli): read macOS tool paths the same way on every host
Follow-ups from review on #7443 and #7444:
- The quote hint returned the first run of words that existed, so a
"Final Cut/" folder hid "Final Cut export.mp4". It now names the longest.
- "--compare=my clip.mp4" got no hint; the value after = now starts a run.
- The telemetry section lists the skill a reverse or templates command ran
for, and the HeyGen session id a recorded reverse run carries.
* feat(cli): reverse and templates report their own runs; commands say which skills are installed
Gaps in what skill telemetry can answer for /reverse-any-video and
/style-sampling:
- The skill's beacon fires only when the agent runs it, and agents often
skip it. `reverse start` and `templates pull` now send `skill_invoked`
with `source: command`; beacons carry `source: beacon`. A person counts
once.
- style-sampling renders were never tagged. A pull marks the project the
episode is built in (only a folder that is already a project).
- Installs of an on-demand skill were invisible. `cli_command` carries
`skills_installed` and `skills_outdated`, the names from the cached
background skills check, so "never triggered" and "not installed" can
be told apart.
- `auth status` signed out exits 1 with a bare "Command failed", which
reads like a login failure. It now carries `NotSignedIn`, and an auth
error its own cause.
And a step that gets lost: a reverse started signed out often goes on
without ever trying to sign in. `reverse start` signed out now prints
`login_command` and a `next` line in the JSON the agent reads, not only
in the skill's reference.
* refactor(cli): keep comment share at the base in the reverse telemetry files
* fix(cli): command runs get their own event, counted once per piece of work
Review on the first cut: `skill_invoked` from the commands doubled the
#7176 beacon's single observation, and it fired before any work.
- `reverse start` and `templates pull` send `skill_command_ran`, not
`skill_invoked`; the beacon is back to its #7176 shape.
- `reverse start` sends it after the probe, with `signed_in` and the
reverse `session_id` that keys a retry; a video it cannot probe sends
nothing.
- `templates pull` sends it only when it downloads; a re-pull or a
refused pull sends nothing.
- A skills check that finds no install forgets the installed names.
- `auth status` failures carry only a code, never the auth error's text.
* fix(cli): measure answers the four mistakes agents make most
The usage errors agents hit most on `measure`: no op named, a path it
cannot read, and a flag it does not take. Some agents never get past
them to a successful measure.
- `measure <video>` with no op (alone or with --json) runs `probe` and
says so on stderr. With other flags the op is a guess, so the
unknown-command hint asks for one; it now names a path that does not
exist yet too.
- When ffprobe cannot read a file, the error carries ffprobe's reason and
a matching fix: a cut-short download, a page saved as .mp4, a
permission.
- An unquoted path with spaces says to quote it, both where the extra
words are rejected and where the cut path is missing.
- Window flags take --from/--to, and folder flags take -o.
* refactor(cli): keep comment share at the base in the measure files
* test(cli): expect a spaced path as the shell split it, on Windows too
* fix(cli): a cut-short MP4 says download it again, read from its header and ffprobe's specific line
* fix(cli): an unquoted spaced path reaches probe or its quote hint wherever it falls
* feat(skills): reverse-any-video and style-sampling install with the core set
A skill only triggers when it is installed. On demand these two rarely
were, so a request to copy a video or make an episode in a proven style
seldom reached them, and the entry skill named only one of them, in its
closing line.
- isCoreSkill, FALLBACK_CORE_SKILLS and the marketplace `core-skills`
entry include both, so `init` and `skills update` install them. The
`hyperframes` plugin already bundled every skill.
- `/hyperframes` routes to both from the top of its table: a video to
copy or remake, and an episode in a proven style. Its budget takes
+64 B for those two rows.
- `/reverse-any-video` triggers on "remake" and on 「复刻这个视频」
「照着做一个」, inside the 320 B description cap.
- Every catalog surface lists both in the core group: CLAUDE.md,
AGENTS.md, README, the skills guide, skill-lifecycle and the two init
templates (still byte-identical).
* refactor(cli): keep comment share at the base in skillsManifest
* feat(producer): distributed and range renders honour page motion blur
A film with data-motion-blur="on" blurs the same way however it is rendered. plan() no longer
forces the synthetic job to pageMotionBlur "off", so the plan locks screenshot capture for a
blurred film as a whole render does, and renderChunk() no longer defaults the page's off switch
to "distributed-render". An explicit pageMotionBlur "off" in the config still stands the page down.
Hosted renders pick this up after the next OSS release and a producer pin bump; a render of a
blur-on film that goes distributed changes output then.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* test(producer): the default tier blurs the same in chunks; docs agree distributed renders blur
Two files compare plan -> renderChunk -> assemble with the tier attribute removed, so `auto`
(the filter tier) decides, against a whole render: held-card and callbacks (the muted-callback
path). The skill's motion reference no longer says distributed renders are off. The rewritten
R19 plan test gets 60 s for its three plan() calls.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
CLAUDE_CODE_ENTRYPOINT is set by current Claude Code versions too, and it
is the only Claude marker MCP servers inherit, so it is not an
"older versions" fallback. The Agent SDK comment still said Claude Code's
shell tool sets CLAUDECODE; the rule keys on CLAUDE_CODE_CHILD_SESSION.
Follow-up to review nits on #7425. Comments only.
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(cli): add agent markers other detectors use, verified upstream
Compared std-env, @vercel/detect-agent, am-i-vibing and agent-cli-detector
against our rules and kept only markers the agent itself sets on the
shells it spawns: Auggie AUGMENT_AGENT, Amp AMP_CURRENT_THREAD_ID, Cursor
CURSOR_CONVERSATION_ID and CURSOR_SANDBOX, Cline CLI CLINE_WRAPPER_PATH,
Muse MUSE_RELEASE_INFO, and Warp Oz OZ_RUN_ID (after the harnesses it
wraps).
Claude Code now keys on CLAUDE_CODE_CHILD_SESSION or
CLAUDE_CODE_ENTRYPOINT. Its IDE extensions also set CLAUDECODE in the
integrated terminal a person types into, so CLAUDECODE alone counted that
person as an agent: no prompts, agent-mode login.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* docs(cli): trim comments and fix citations the comment gates flag
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore: raise the cli packed size budget to its measured size
main already sat at the edge of the 1% tolerance; the new agent rules
put the packed CLI 842 bytes past it (8062664 vs 7982002 + 1%).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
TERM_PROGRAM=cursor|windsurf and Replit's REPL_ID only say where a shell
runs. They now sit in their own list, checked after the AI_AGENT / AGENT
name check, so Amp in a Cursor terminal or Devin in Replit report as
themselves, and an unknown AI_AGENT there keeps its discovery hints.
agentOnly had no users left, so it is gone.
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(cli): detect more agent runtimes in telemetry
Agent runtime rules:
- copilot_agent: key on COPILOT_AGENT=1 or AI_AGENT=github_copilot_*
(VS Code agent terminals) instead of the unverified Actions markers.
- copilot_cli: COPILOT_CLI=1 or COPILOT_AGENT_SESSION_ID, ahead of
copilot_agent.
- kilo (KILO=1, KILOCODE_VERSION), ahead of opencode.
- opencode (OPENCODE=1, OPENCODE_SESSION_ID).
- qwen_code (QWEN_CODE=1), junie (JUNIE_SHIM_PATH),
antigravity (ANTIGRAVITY_AGENT), qoder (QODER_AGENT),
kiro (AGENT_DISPLAY_OUT, AGENT_CONTEXT_OUT), openhands (AI_AGENT=openhands).
- AI_AGENT fallback after every vendor rule: known names map to their
runtime, any other slug-safe name becomes other_ai_agent. Bare AGENT
is not used.
Discovery hints: other_ai_agent keeps its hints, agent_hint prefers
AI_AGENT over AGENT, and AGENT_TOOLSDIRECTORY,
GSM_SKIP_SSH_AGENT_WORKAROUND and EDGE_METRICS_SESSION_ID are no
longer hint keys.
CI: name AWS CodeBuild (CODEBUILD_BUILD_ID) as codebuild.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(cli): detect agents seen in discovery hints
Add rules for closed-source agents whose markers show up in the CLI's
own agent_env_hints on non-interactive runs: WorkBuddy, CodeBuddy, Trae,
Pippit, Doubao, Aily, Grok, Kimi, ZCode, SWE-agent and apps built on the
Claude Agent SDK. Qoder also matches its CLI and agent SDK keys.
The AI_AGENT name map now covers the names agents report (Devin, Amp,
v0, Rovo Dev, Base44, Trae, Antigravity, OpenHands, Copilot, and
hermes-agent), and a known name in the bare AGENT variable counts too.
Unknown AGENT values such as AGENT=1 still do not.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(cli): detect the DeepSeek harness
Every dsh model shell call runs with DSH_SHELL=1 and the agent's
DSH_SESSION_ID. HyperFrames Desktop also runs dsh; its runs carry
HYPERFRAMES_CLIENT, so the client field separates the two.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* docs(cli): cite the Agent SDK source for claude_agent_sdk
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(cli): let the closest agent win over inherited markers
Copilot's COPILOT_AGENT / AI_AGENT and COPILOT_CLI pass down to the
agents it starts, and TERM_PROGRAM / REPL_ID only say where a shell runs.
Those rules now come after every other agent rule, so opencode or Qwen
started by Copilot, or Kilo in a Cursor terminal, report as themselves.
dsh keys on DSH_SHELL=1 only: it sets DSH_SESSION_ID on the terminal a
person types into too. Qoder drops its app-wide SDK and setting keys,
CodeBuddy drops CODEBUDDY_AGENT_RUNTIME (a setting it reads), and Kiro's
markers must be non-empty.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat(cli): add 1440p resolution presets to --resolution
Adds landscape-1440p (2560x1440), portrait-1440p (1440x2560) and
square-1440p (1440x1440) next to the existing 4K, 720p and 480p tiers,
with the aliases 1440p, 2k and qhd plus 1440p-portrait and 1440p-square.
1440p joins the aspect-agnostic aliases, like 720p and 480p.
Every validator derives from CANVAS_DIMENSIONS, so no logic changes: a 4K
composition renders down to 1440p through the existing smaller-preset
path, and a 1280x720 composition supersamples to 1440p at an exact 2x.
A 1920x1080 composition is still refused by the non-integer-scale check
(4/3); fractional scale factors are not part of this change.
Refs #6309
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(cli): clarify 1440p aliases and repair skill validation
---------
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
Co-authored-by: James <james.russo@heygen.com>
* fix(cli): cloud render finds media whose file name is stored in another Unicode form than the page's
* fix(cli): keep the archive-key change from raising publishProject's comment share
* fix(cli): a page that names a decomposed file as the disk lists it still matches its archive entry
* feat(cli): check returns an unchanged project's last pass in about a second
* fix(cli): the check cache hashes files without a stat race and its tests run before the signal test
* test(cli): the warm check test compares two real runs, not a run and its cached pass
* feat(cli): check --no-craft skips the craft pass for a render preflight
* fix(cli): the check cache keys on timeout, env and large files and never replays a moved tree
* style(cli): format the check cache
* fix(cli): the large-file cache test waits for the clock tick it depends on
* fix(cli): the check cache keys on the browser and on scripts under .hyperframes
* fix(cli): the check cache ignores the video copies the check itself touches
* fix(cli): a media file referenced by its filesystem path goes into the cloud archive
* fix(cli): an absolute media path keeps its fragment and query, and the resolver is split
* fix(cli): a Windows drive-letter media path is a filesystem path too
* fix(check): lint and check skip compositions the project never mounts
* refactor(lint): let two new names stand without doc comments
* test(cli): lint tests that lint unmounted files say so
* refactor(cli): the cloud upload's PUT and retry code moves to its own file
* feat(cli): cloud uploads send a big file in parts, so a cut connection resends one part
* test(cli): the cutting link fails a single PUT of the same file
* fix(cli): a file the cloud refuses forgets its saved parts plan
* fix(cli): a resumed upload goes straight to its parts, so a saved plan never feeds the single PUT