Commit Graph
1405 Commits
Author SHA1 Message Date
Miguel Ángel e3ff9c803c fix(engine): start nvidia-smi, vm_stat, ps and tasklist from trusted folders only (#7788)
* fix(engine): start nvidia-smi, vm_stat, ps and tasklist from trusted folders only

The renderer resolves these helpers through the shared trusted lookup and runs the absolute path, so a program with the same name in a project folder is never started in their place.

* test(engine): memory and import tests follow helpers started by resolved path
2026-11-14 20:14:27 -05:00
Miguel Ángel 5e0577f124 chore: release v0.9.50 2026-11-14 19:39:50 -05:00
Miguel Ángel f4b6aad332 fix(engine): media playlists are checked through every list they name (#7778)
* fix(engine): media playlists are checked through every list they name

* refactor(engine): split the nested playlist walk into small steps
2026-11-14 19:35:58 -05:00
Miguel Ángel 06cf964646 fix(engine): refuse media playlists that name files outside their folder (#7768)
* fix(engine): refuse media playlists that name files outside their folder

* refactor(producer): media preflight decides refusals in one helper

* fix(engine): read a playlist once, bounded, instead of sizing it first
2026-11-14 18:54:57 -05:00
Michael WangandClaude Opus 5.5 603395c89a fix: render file servers refuse env and Git files; shown-box sampling stays within the render (#7765)
* fix(producer): shown-box sampling stays within the validated render duration

Sampling where each oversized video is drawn stepped through the authored
start and end in 0.25 s float steps with no bound, before the render
duration was checked. An enormous start never advanced the step, and an
unbounded end grew the sample set until memory ran out.

The render duration is now validated first, each window is cut at the
composition's end, a window that is not finite is skipped, and the loop
counts integer steps. Building the sample list happens inside the
measurement's try, so any failure falls back to full-size extraction.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix: render and capture file servers refuse env files and Git data

The producer render file server, the engine file server and the CLI
static project server served every regular file under the project to the
page, including `.env*` files and anything under `.git`. Compile already
refuses to inline those files, and Studio preview and play refuse to
serve them.

All three now answer 404 for them through the shared isVcsOrSecretPath
rule, checked on the requested name and on a link's target, with the
case and Windows spelling variants it already covers.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-14 23:36:46 +00:00
Michael WangandClaude Opus 5.5 0e767d620f fix(producer): render-service media probes stay on public HTTPS (#7764)
* fix(producer): inline only public HTTPS scripts, under the download URL policy

The compile step fetched every external `<script src>` with a bare fetch:
any scheme, any host, redirects followed, body size unbounded.

It now fetches through the engine's download policy, the one remote
stylesheets and media already use: HTTPS only, literal and resolved
addresses public, every redirect hop checked again, connect pinned to the
checked address, and at most 16 MiB per script. A new engine helper,
fetchPublicHttpsBytes, returns the raw bytes so integrity metadata is still
checked before decoding; fetchPublicHttpsText now decodes its result.

A script outside the policy keeps its `<script src>` tag, and the capture
browser loads it as before; the log says so at info level instead of
warning that the render may fail. Scripts under the operator's
HYPERFRAMES_GSAP_BASE_URL mirror, which may be loopback HTTP, are still
inlined. An integrity mismatch still fails the render.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(producer): the audio probe skips remote video URLs outside the public HTTPS policy

A video whose remote src compile could not download keeps that URL, and
the probe-only audio pass handed it straight to ffprobe, which fetches any
scheme and host. A src that fails the literal public-HTTPS check (plain
HTTP, loopback, private, link-local or `.internal` hosts) is now skipped,
the same outcome frame extraction already gives it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(producer): the GSAP mirror exception covers only its dist folder

Scripts from the operator's HYPERFRAMES_GSAP_BASE_URL mirror are fetched
outside the public-HTTPS policy, because the mirror may be loopback HTTP.
That exception now covers only plain script file names in the GSAP dist
folder the compiler itself points at, refuses redirects and stops at
16 MiB, like every other inlined script. Any other path on the mirror's
origin follows the public-HTTPS policy and is left for the browser.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(engine): on a render service, https media inputs open only HTTPS

ffmpeg and ffprobe read some https sources in place. An https playlist
or manifest can name segments on any protocol, and the default protocol
list for an https input does not restrict them. With
PRODUCER_RESTRICT_PAGE_NETWORK=true, these inputs now pass
`-protocol_whitelist https,tls,tcp`, so segments and parts they name must
be HTTPS too. Local renders are unchanged.

A new test pins that local project files which are playlists, concat lists
or DASH manifests (with their own extension or named `.mp4`) are probed
and decoded without any request to the URLs they list: ffmpeg limits what
a local file input may open to file, crypto and data, and the concat
demuxer refuses URL entries in safe mode.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-14 23:25:18 +00:00
Miguel Ángel fd7217c8af chore: release v0.9.49 (#7767) 2026-11-14 18:15:46 -05:00
Miguel Ángel cd55f059c3 fix(producer): compile downloads composition scripts through the shared asset downloader (#7763)
* fix(producer): compile downloads composition scripts through the shared asset downloader

* fix(producer): compile downloads composition scripts through the shared asset downloader

* fix(producer): compile downloads composition scripts through the shared asset downloader
2026-11-14 18:01:01 -05:00
Miguel Ángel 89bca002ef feat(engine): renders log the class of every address the proxy refuses (#7762) 2026-11-14 17:30:51 -05:00
Miguel Ángel f33f137b2d chore: release v0.9.48 (#7761) 2026-11-14 17:04:22 -05:00
Miguel Ángel a6ff3922b4 fix(engine): renders reach only public addresses from every part of the browser (#7758)
* fix(engine): renders reach only public addresses from every part of the browser

* fix(engine): renders reach only public addresses from every part of the browser

* fix(engine): renders reach only public addresses from every part of the browser

* fix(engine): renders reach only public addresses from every part of the browser
2026-11-14 16:38:35 -05:00
Miguel Ángel 5aea6d5b68 Merge pull request #7754 from heygen-com/release/v0.9.47
chore: release v0.9.47
2026-11-14 14:48:31 -05:00
Miguel Ángel b100cf4615 fix(engine): look a host up again after a failed lookup instead of refusing it for the whole page (#7751)
* fix(engine): look a host up again after a failed lookup instead of refusing it for the whole page

* fix(engine): name the failed-lookup reset instead of commenting it
2026-11-14 17:24:42 +00:00
Miguel Angel Simon Sierra 2f46a88dea chore: release v0.9.47 2026-11-14 12:13:11 -05:00
Miguel Ángel 29533f631f fix(engine): hold the render page to the asset download address rules (#7748)
* fix(engine): hold the render page to the asset download address rules

With PRODUCER_RESTRICT_PAGE_NETWORK=true the render page reaches only its own project server and public addresses, checked by host and by resolved address, the rule asset downloads already follow. Unset, nothing changes for local renders.

* fix(engine): keep the page network filter comment within four lines
2026-11-14 16:38:55 +00:00
Miguel Ángel b746a0ee13 fix: check, render and downloads work on a drive without hard links (#7737)
* fix(studio-server): check and render keep their history on a drive without hard links

The history lock had its own list of the errors that mean a drive cannot
hard-link, and it lacked EISDIR, which Windows returns on a Google Drive
folder. Every check report and render record there failed. The lock now
uses core's list, the one owner, which also gains ENOSYS from the lock.

* fix(engine): downloads and plan blobs publish on a drive without hard links

The producer's plan publisher kept a third list of no-hard-link errors
without EISDIR, and the downloader had no fallback, so both failed on a
Google Drive folder. Both now use core's cannotHardLink; the downloader
copies into an exclusively created file, so it still never overwrites.

* fix(engine): a download on a drive without hard links publishes atomically

Instead of copying into an exclusively created file, the fallback renames
the staged download into place under the cache path lock when nothing is
there yet, so a killed run never leaves a truncated file to reuse.
2026-11-14 15:47:18 +00:00
Miguel Ángel 43132411c8 chore: release v0.9.46 (#7732) 2026-11-14 13:40:20 +00:00
Miguel Ángel ff9c3b7619 fix(engine): an unrecognised capture failure is unknown, not the author's fault (#7730)
* fix(engine): an unrecognised capture failure is unknown, not the author's fault

* fix(engine): carry the extractor's disk-full words in a name, not a comment
2026-11-14 10:28:28 +00:00
Miguel Ángel fe8d93056c chore: release v0.9.45 (#7729) 2026-11-14 09:45:48 +00:00
Miguel Ángel 58ef24c6d8 chore: release v0.9.44 (#7708) 2026-11-14 06:22:47 +00:00
Miguel Ángel 798c241e4d feat(cli): name a caller's owner id on the browsers and warm check it starts (#7687)
A caller that runs several renders at once sets HYPERFRAMES_OWNER to a per-run id; every browser the CLI or
engine launches and the warm check server carry --hyperframes-owner=<id>, so the caller can tell one run's
leftovers from another's.
2026-11-14 02:42:11 +00:00
Miguel Ángel c727e890f2 chore: release v0.9.43 (#7664) 2026-11-13 22:42:53 +00:00
Miguel Ángel 273672ab13 fix(engine): render text without ClearType colour fringes on Windows (#7663) 2026-11-13 20:55:10 +00:00
ClippyandClaude Opus 5.5 f01209d4ae fix(deps): bump tar, hono, @hono/node-server, adm-zip, sharp and postcss-selector-parser (#7656)
Raise the direct ranges so the workspace and published packages resolve
patched releases: tar 7.5.22, hono 4.13.16, @hono/node-server 2.1.6,
adm-zip 0.6.1, sharp 0.35.5 and postcss-selector-parser 7.1.7. The CLI's
on-demand sharp install moves to 0.35.5 with them.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:18:17 +00:00
Miguel Ángel 30c0935a24 chore: release v0.9.42 (#7655) 2026-11-13 17:54:21 +00:00
ClippyandClaude Opus 5.5 7603215f03 chore(deps): override tinypool to 2.1.2, bump vite and vitest patches (#7653)
oxfmt 0.41 pins tinypool 2.1.0 (GHSA-5gmw-xhrv-c9v3, GHSA-85c8-ppgw-ccpr).
Every oxfmt release before 0.67 pins 2.1.0, and 0.67+ reformats 66 files,
so override tinypool to the drop-in 2.1.2 patch instead of bumping oxfmt.

Also bump vite ^6.4.2 -> ^6.4.5 and vitest/@vitest/coverage-v8
^4.1.11 -> ^4.1.12 (pulls @vitest/mocker 4.1.12).

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 17:30:24 +00:00
ClippyandClaude Opus 5.5 9d1e4e54a7 fix(deps): bump puppeteer to 25.13 so chromium-bidi resolves zod 4.6.7 (#7652)
puppeteer-core 25.13.0 moves chromium-bidi to 157.0.8090-0, which depends on
zod ^4.6.5 instead of ^3.24.1. Lifts the engine/producer puppeteer pin from
~25.11.0 to ~25.13.0, keeps one puppeteer-core and one @puppeteer/browsers
(3.2.4) across the workspace, and updates both Lambda runtime locks with a
targeted npm resolution so only the puppeteer subtree changes.

Clears GHSA-mrhx-c7xc-m3w5 in both runtime locks.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 17:30:21 +00:00
renovate[bot] d79e72c024 fix(deps): update dependency undici to v7.30.1 [security] (#6695)
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2026-11-13 15:31:55 +00:00
Miguel Ángel 20084cbe67 chore: release v0.9.41 (#7632) 2026-11-13 04:26:22 +00:00
Miguel Ángel b54e4981f2 fix(engine): keep media ids that differ only by case in separate folders (#7635)
* fix(engine): keep media ids that differ only by case in separate folders

macOS and Windows disks ignore case, and APFS ignores Unicode normalization,
but mediaIdPathSegment passed plain ids through as written. Two clips such as
`s12-vB1` and `s12-vb1` therefore shared one `__hyperframes_video_frames`
folder: whichever was staged first kept it, and the other clip read its
frames. A longer clip then failed mid-render with ENOENT on a frame past the
shorter clip's end; a same-length pair rendered the wrong footage silently.

Only a lowercase, NFC-normalized id now maps to itself. Any other spelling
gets the existing readable~digest form, so every per-id path (frame folders,
the tail probe, audio work files, distributed plans) stays distinct on a
case-insensitive disk.

* fix(engine): name the case-folding check instead of commenting it

Keeps mediaPathSegment.ts at or under its comment share at the merge base.
2026-11-13 03:09:18 +00:00
James RussoandClaude Opus 5.5 9701f4bbf3 fix(engine): diagnose a failed Chrome launch with the env it had (#7634)
The startup diagnosis rebuilt its env from process.env, so a launch that
dropped WAYLAND_DISPLAY or the SwiftShader Vulkan pins was diagnosed with
them restored. The engine and CLI now pass the failed attempt's env through
explainChromeLaunchFailure; process.env stays the fallback when no env is given.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 01:36:12 +00:00
James RussoandClaude Opus 5.5 24d56e6813 fix(engine): launch Chrome with only the env it needs (#7629)
* fix(engine): launch Chrome with only the env it needs

Every Chrome HyperFrames starts (render, GPU probe, HDR capture, startup
diagnosis, version and shell probes, CLI browsers, inspector) now gets an
allowlisted environment from @hyperframes/core/chrome-env instead of the
whole process env. HYPERFRAMES_CHROME_ENV_PASSTHROUGH names any extra
variable a host needs.

Refs #7623

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(core): keep the Chrome env passthrough across repeat filtering

The CLI GPU probe filters its env twice, which dropped the names
HYPERFRAMES_CHROME_ENV_PASSTHROUGH adds, so the probe and the render
saw different envs. Keep the passthrough variable itself.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 00:53:11 +00:00
James RussoandClaude Opus 5.5 43722355a3 fix(engine): diagnose the launch attempt whose error surfaces (#7630)
When a sandboxed Linux launch fails and the no-sandbox retry also fails,
the startup diagnosis re-ran Chrome with the first attempt's flags, so it
described a launch that was not the one reported. The engine and CLI now
diagnose the flags of the attempt that produced the error.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 00:34:30 +00:00
James RussoandClaude Opus 5.5 400a4e2285 fix(engine): keep Chrome's sandbox on for local renders where the host supports it (#7625)
* fix(engine): keep Chrome's sandbox on for local renders where the host supports it

Chrome launched with --no-sandbox on every host. Keep the sandbox on
macOS, Windows and non-root Linux outside containers; drop it as root,
in Docker/Podman/Kubernetes, on Lambda and Cloud Run, and where Linux
restricts unprivileged user namespaces. --no-zygote now ships only with
--no-sandbox, since Chrome refuses it while sandboxed. A sandboxed Linux
launch that fails at startup retries once without the sandbox and warns.
HYPERFRAMES_CHROME_SANDBOX=1|0 overrides the choice.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* test(cli): cover the Linux sandbox retry in managed browser launches

The off-macOS closed-pipe case now passes --no-sandbox, so it still
pins "no seatbelt relaunch" rather than the new Linux sandbox retry,
which gets its own case.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(engine): honor HYPERFRAMES_CHROME_SANDBOX=1 and latch the fallback on clear evidence only

With the override set to 1, a sandboxed launch Chrome refuses now fails
instead of retrying unsandboxed. A bare startup exit (TargetCloseError)
retries that launch without the sandbox but only latches the sandbox off
after a second one, so a single unrelated early kill no longer disables
it for the rest of a long-lived process. Chrome naming the sandbox in its
error still latches at once.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 23:16:45 +00:00
Miguel Ángel 9c3258ccbf fix(engine): fast capture works on Chrome 157+, which wants content=drawable (#7626) 2026-11-12 23:01:15 +00:00
Vance IngallsandClaude Opus 5.5 2df837be9e test(engine): the blur trail mask covers the shutter's sweep, not only whole frames (#7624)
A mover that moves further than its own width between frames leaves gaps in a
trail built from whole-frame blur-off poses; the shutter's smear lands in them and
read as a leak outside the trail (nightly seed 63098). The mask now adds the
shutter's sub-frame poses, captured in time order. Seed 63098 joins the default
seeds as the regression case.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 22:51:07 +00:00
Miguel Ángel f5476851dd chore: release v0.9.40 (#7617) 2026-11-12 21:52:46 +00:00
Miguel Ángel 97df93fb07 fix(cli): check hears a WAV whose header misstates its byte rate, as render plays it (#7618)
* fix(cli): check hears a WAV whose header misstates its byte rate, as render plays it

A PCM WAV whose header byte rate is wrong (some speech generators write 2x) sent the silence decode's input seek past the file's end, so every clip past the middle was reported audio_clip_silent. Render already cuts such a file by decoded time (#7541); the byte-rate rule now lives in parsers and the CLI decoder seeks after decoding starts for those files.

* test(cli): keep the misstated byte-rate case with the other single-clip checks

* style(cli): format the WAV test helper

* style(parsers): keep the byte-rate rule's comment within the line limit
2026-11-12 19:30:02 +00:00
Märten Tropp b4fc4736d7 fix(engine): a frame's pixels no longer depend on the frames drawn before it (#7181) 2026-11-12 17:52:38 +00:00
Miguel Ángel 54bcdc616f fix(producer): a local render extracts each video at about the size it is shown (#7609) 2026-11-12 17:16:59 +00:00
Miguel Ángel 395c845cd3 fix(engine): render work is sized by the CPUs the process may use (#7612)
* fix(engine): render work is sized by the CPUs the process may use

* fix(engine): keep the usable CPU count comment under the line limit
2026-11-12 17:07:45 +00:00
Miguel Ángel cc47f7ea74 fix(engine): a video shown at exactly half its size is extracted at half size (#7607) 2026-11-12 14:00:10 +00:00
Miguel Ángel 256675f92c chore: release v0.9.39 (#7605) 2026-11-12 13:27:37 +00:00
Miguel Ángel 1ec03c4e19 chore: release v0.9.38 (#7594)
* fix(scripts): release:prepare refuses a baseline that skips a release whose tag this clone lacks

* chore: release v0.9.38
2026-11-12 08:29:06 +00:00
terencechoandClaude Opus 5.5 e8e1177ad6 chore: release v0.9.37 (#7578)
* chore: release v0.9.37

* docs(release): add #7576 to v0.9.37 notes

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 02:01:41 +00:00
Vance Ingalls f66c673d19 chore: release v0.9.36 (#7575) 2026-11-11 16:06:36 -08:00
terencechoandClaude Opus 5.5 6bd9bb8a44 chore: release v0.9.35 (#7571)
* chore: release v0.9.35

* docs(release): note the render stop and container 403 from #7563 in v0.9.35

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-11 22:09:19 +00:00
Miguel Ángel 7a4dfb7654 chore: release v0.9.34 (#7555) 2026-11-11 16:46:12 +00:00
James RussoandClaude Opus 5.5 84799bdc59 fix: take helper programs only from install locations, never the project folder (#7524)
* fix(cli,parsers): take helper programs only from install locations, never the project folder

FFmpeg, FFprobe, whisper-cli, parakeet-mlx, Python and system Chrome are now
found through one shared rule: an absolute PATH entry outside the working
directory, or a fixed install dir. A relative PATH entry and any folder
inside the working directory are skipped, on every platform. Windows walks
PATH itself instead of asking `where`, which lists the working directory
first.

FFmpeg discovery no longer falls back to `.hyperframes/bin` or to an
installer package in the project's node_modules; an installer package
installed next to HyperFrames still works. The env overrides are unchanged.
The CLI skill reference stops suggesting the project-local folder.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* test(cli): give the PATH-lookup tests a fixed PATH on Windows

Windows now walks PATH instead of running `where`, so tests that faked
`where` output found the runner's real python and cmake. Both tests now set
PATH to a temp folder holding only the programs they expect.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix: report a missing helper program instead of running it by bare name

When no trusted FFmpeg or FFprobe is found, the engine, the producer's
probes and Studio's waveform, peak and audio-track helpers now report
"not found" with the HYPERFRAMES_FFMPEG_PATH / HYPERFRAMES_FFPROBE_PATH hint
instead of handing the bare name to the OS lookup. Callers that degraded
on a missing binary still do: runFfmpeg returns a spawn_error result, the
GPU and version probes report none, and Studio's audio helpers fail soft.

The Lambda handler names its bundled ffmpeg and ffprobe by path, since
they live under the task root, which is the working directory.

brew, git and cmake for the whisper.cpp install, and the clipboard tool
for `hyperframes add`, run by the absolute path the trusted lookup
approved. The media-use engine gets a PATH holding only trusted folders.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* fix(engine): fail a missing FFmpeg the way a failed spawn always has

Throwing from getFfmpegBinary turned every caller's missing-binary path
into a synchronous throw: the streaming encoder's "failure result, not a
throw" contract broke, and unit lanes that run without FFmpeg crashed at
module load. With no trusted copy the getters now return an absolute path
inside HyperFrames' own install that is never created, named for the
override to set. Spawning it fails with ENOENT exactly as before, and the
OS still never looks the bare name up. The call-site guards and fixture
changes that the throw needed are reverted.

The whisper install tests match git and cmake by name and skip the brew
case on Windows, where the lookup walks the real PATH.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* test(engine): pin the found FFmpeg before the HDR test looks it up as macOS on Windows

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-11 09:46:52 +00:00
Miguel Ángel f312fca7e2 feat(producer): fail a render whose variables the page does not declare (#7548)
* feat(producer): fail a render whose variables the page does not declare

* chore(producer): keep the comment share and the dead-code audit clean for the variables check
2026-11-11 08:46:13 +00:00