| Files | Last commit | Last update |
|---|---|---|
fix: 1.17.1 — fit-size reachable from templates, SKILL.md routes the 1.17 features, UTF-8 child output (#234) (#237) * fix(render): fit the caption size on the template path The caption size a template carries comes from the platform table, not from anyone stating it, so forwarding it as a bare --size switched off caption.py --fit-size auto and long cues were split instead of shrunk (eval 18 cw1). Templates now default the captions block to fit_size "on" (a template or a brand.json that states a size still wins), project.json accepts fit_size/min_size/fit_size_scope, and render.py reports the caption stage stats. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs(skill): route filler, beats, --jobs and --cache; one label for a partial result Eval 18 showed agents could not discover the 1.17 features: SKILL.md never mentioned filler, --snap beats, --jobs or --cache. Adds one routing row each (paying for them with duplicated wording elsewhere) and states that a partial result is Done: with the shortfall in Notes:, never a third label. Also: caption.py reports text_unchanged, and evals/write_fixtures.py stages a batch recipe output_dir absolute under OUTDIR. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: 1.17.1 changelog, roadmap, contract rows and render/caption reference Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs(readme): describe the template caption path, fit-size reachability and the one report label Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * test(caption): assert the text-unchanged disclosure Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs(skill): keep the -o consent rule, drop the no-action dimensions line Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix(runner): decode child output as UTF-8, not the locale code page (#234) On a Windows cp932 machine text=True decoded ffprobe JSON with the console code page: the reader thread raised UnicodeDecodeError, communicate() returned an empty stdout, and probe.py reported "?s | no video | no audio" with exit 0. Every child capture now states encoding="utf-8", errors="replace", and a probe with no ffprobe output refuses (kind: input) instead of reporting nulls. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * evals: ml2 tests the subtitle mux, not a translation The prompt asked for a translation the skill must refuse, so the run never reached the muxing path it was meant to exercise. It now ships de.srt and a video and expects caption|render. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: record the 2026-09-14 external review, the Japanese boilerplate example and a cache failure-path test SKILL.md now shows the Japanese rendering of the Check:/Look: filler lines; docs/design-decisions.md records accept/reject for each review item; the release checklist pins the roadmap line to the bump PR; and the --cache failure path (nothing cached, no work directory left) is now tested. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix(waveform): drop the duplicated encoding keyword Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix(render): fit the caption size unless the brand states one `--brand` alone is not a stated caption size: brand_caption_style() applies no defaults over what the file said, so a brand of colours or a font left the platform table's 24 in place while withholding fit_size -- exactly the 1.17.0 behaviour 1.17.1 exists to fix. Only a brand file that actually names caption.size (checked against `_stated`, since BRAND_DEFAULTS always supplies one) now stands the fitter down. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix(runner): decode measurement output as UTF-8 too (#234) run_analysis() is the one child capture the #234 pass missed, and it is the measurement path: loudness.py and check.py parse the loudnorm JSON out of its stderr, as do cropdetect, stabilize, probe and asr. With the locale's codec a UTF-8 filename echoed by ffmpeg raised UnicodeDecodeError inside subprocess.run -- an unhandled traceback, not a `status: failed` document. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * test(contract): scan the parsed call, and evals/, for locale-decoded captures The #234 guard matched only lines containing `text=True`, so run_analysis()'s `text=text` was invisible to the test whose whole job is to catch it. It now parses each subprocess.run/Popen call and flags any text-mode capture that does not name UTF-8, with text=False and a **mapping of encoding kwargs as the only exemptions, and it scans evals/ as well -- where two captures (contract/check.py, grade_runs_24.py) still decoded with the locale. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * test(contract): make the #234 probe tests state which sh() they call Review 17 finding 4 reads these as calling tests/_fixtures.sh (no `check` parameter) and so never asserting. They call this module's own sh() at line 41, which does take check=, and both tests run and pass. The success case now relies on sh()'s own check instead of a redundant kwarg, and the refusal case says in a comment which helper it uses so the next reader is not misled. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix(caption): text_unchanged is false when a cue was split or a glyph stripped The key claimed the drawn text equalled the cues handed in on runs that split a cue across two consecutive cues -- the very defect 1.17.1 exists to fix -- and on `--emoji none`, which replaces the clusters out of the drawn text. Wrapping, line breaks and timing still do not count: the words are the same. docs/contract.md and references/scripts.md now state the rule, and that the key is burn mode only (`--mode mux` never touches the text and omits it). Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs(skill): restore the dimensions-and-rotation rule, pay for it in wording Review 17 finding 6: the yuv420p/rotation bullet was cut outright, and it was the only link anywhere to references/gotchas.md#dimensions-and-rotation, so that section was unreachable from the agent's entry point. It is back in one clause. (The `doctor --json` `fonts.scripts` pointer is not lost -- it survives in the CJK bullet of the judgement-call list.) Paid for by trimming wording: the reframing bullet's fps clause repeats the reframing gotcha, and five other phrases lost redundant words. 29,982 bytes, budget 30,000. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: 1.17.1 is planned for release, and the released version is 1.17.0 The three version documents disagreed: package.json says 1.17.0, the CHANGELOG files 1.17.1 under Unreleased (where the release workflow expects it), and the roadmap already called 1.17.1 shipped. The roadmap row is now `planned` -- the state 1.16.1's row carried before its tag -- and says it is not tagged yet. The "released version today" line, which still read 1.16.0 against a 1.17.0 package.json, now reads 1.17.0, and design-decisions stops recording P0-3 as closed by a PR that only landed the checklist half of it. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs(contract): name both cases where the caption block is null `caption` is also null when the captions stage was served from the --cache (a cache hit clears the last child document), and `caption.py --mode mux` writes no caption block at all -- neither was stated. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix(evals): judge a staged output_dir absolute on Windows too Path("/srv/out").is_absolute() is False on Windows, so a posix-absolute output_dir in a prompt fixture would be rewritten under OUTDIR there. The check now normalises and asks posixpath/ntpath, and the test asserts both spellings instead of only the posix one. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs(skill): the beats row carries the measured grid to the cut As written the row had cut.py re-measure the grid from its own input -- on a talking-head clip with a separate music bed that is a refusal ("no reliable beat grid in this audio"), not a grid. The row now writes the measurement (`scenes.py track.mp4 --beats --json > beats.json`) and passes it (`--snap-source beats.json`); both commands verified end to end. Paid for within the 30,000-byte budget by trimming wording (29,991 bytes). Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs(changelog): state the 1.17.1 behaviour as it now is The Unreleased entries described the earlier shape of three fixes: a brand file only stands the fitter down when it states caption.size, text_unchanged is false on a split cue or a stripped glyph (and is burn mode only), and the UTF-8 decoding now covers run_analysis() and evals/. Still under Unreleased -- the release workflow moves it. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs(roadmap): one 1.17.1 section after merging main's eval-18 text Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * test(evals): compare the staged output_dir against the resolved tempdir (macOS /private/var) Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs --------- Co-authored-by: Claude <noreply@anthropic.com> | 2 months ago | |
docs: SKILL.md restructured, no behavior change (#392) * Add join.py --copy (stream-copy join of matching clips) and cut.py --remove join.py --copy --transition none joins clips by the concat demuxer with -c copy when their first video and audio streams match (cut.py's stream signature, now in _common.probe), no clip hides edit-list pre-roll, and each clip's picture and sound end within 50 ms. A mismatch is refused by name (join_copy_mismatch). The join is measured beside the output -- video packet count and length against the clips' -- before it is renamed into place. cut.py --remove A-B,... is the inverse of --segments: the ranges are dropped and the rest is cut exactly as the equivalent --segments would be. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Add output_identity, check.py --expect-identity and check.py --compare Every writing tool reports output_identity {size, mtime_ns} for the file it wrote; check.py --expect-identity SIZE:MTIME_NS FAILs an identity row when the file at the path is another one (an older file a failed run left, or a later run's), and check.py reports input_identity. SKILL.md's delivery check passes it. check.py --compare REF measures SSIM (mean and worst frame) and PSNR against a reference in one pass, WARN under a mean SSIM of 0.98. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Add export.py --max-mb: a two-pass x264 encode that fits a size, measured The preset's x264 line takes the video bitrate the running time leaves after the audio (97% of the budget) in place of its CRF and runs in two passes; the written size is a file_size verification step, with a size_over_limit warning when the encoder overshot. Refused for non-x264 presets and --hw, with --crf, without a duration, and under 100 kb/s for the picture. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Add export.py --preset copy --repair: remux a damaged file and count the damage Timestamps regenerated, corrupt packets dropped, demuxing past errors, negative timestamps shifted to zero; the input is only read. Both files are decoded once at -v warning and the timestamp warnings and other errors counted (repair: {before, after}); the repair step fails if either grew, and errors a remux cannot fix are a repair_incomplete warning. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * export.py --repair: judge the remux on timestamp warnings Decode errors are log lines whose count changes with the container (one full run logged 51 before and 53 after), so they are reported before and after but no longer fail the verification step; timestamp warnings, the part a remux rewrites, do. The test source is encoded single-threaded so its damage is the same every run. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Label the join copy, identity, compare, max-mb and repair additions 2.24 2.23.0 shipped with check.py's flashes and first-frame rows; these land in the next minor. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Address review: cover art, timeline times, copy-join guards, --remove edges, --max-mb - check.py --compare reads V:0 (not cover art), times the worst frame by its presentation time, counts frames from packets when nb_frames is missing, prints that note in plain text, runs under --dry-run when both files exist, and starts each main() with no leftover notes. - join.py --copy compares the first non-cover video stream, judges A/V alignment on each stream's end (start + length), includes colour matrix and range in the signature, refuses --sample-rate/--channels, fails the packet check when nothing was counted, and reports short and silent clips as the re-encoding join does. - cut.py --remove refuses any explicit --start and --snap beats, and clamps a range that starts before 0. - export.py --max-mb refuses nan and inf. - contract: join is conditional on audio; check notes name --compare. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Add a cut.py --remove demo; damage only video payload in the repair test The repair test overwrote bytes anywhere in the middle of the TS, which on ffmpeg 5.1.1 hit data its ffprobe could not open the file past; it now flips one byte in 40 video payload packets, never a TS header, the PAT/PMT or audio. The demo shows --remove 2-4 on the gallery's clip. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: refuse a different stream layout and a start offset; --compare pairs by index - The concat demuxer maps each clip's packets by stream index, so a clip whose streams are laid out differently (a cover picture between video and sound) is refused before the signatures are compared. - Picture and sound must start, as well as end, within 50 ms of each other in each clip. - check.py --compare re-times both pictures to their frame index before ssim/psnr, which otherwise pair frames by timestamp. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * check.py --expect-identity: grade an existing file under --dry-run too The identity was skipped for every dry run; it is skipped only for a pending placeholder now, as the content rows are. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy checks starts without stream lengths; --expect-identity reads twice - Matroska/WebM state no stream length, so the A/V guard compared nothing there; the starts are now compared whenever both streams exist and the ends only where both lengths are known. - check.py reads the input's identity before measuring and again after: a file replaced mid-check FAILs, and a path that holds no regular file is a FAIL row rather than no row. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * check.py: read the identity after --compare; note a mid-check change without --expect-identity The second identity read now follows every measurement, --compare's pass included, so a replacement during that pass is caught. Without --expect-identity no row is added (the row set is unchanged), but a file that changed while it was being checked is named in notes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Route the 2.24 features from SKILL.md; make the repair test open on every ffprobe SKILL.md: routing rows for export --max-mb, cut --remove, join --copy, export --preset copy --repair and check --compare; the report format moved above the routing table (inside the first ~5000 tokens kept after compaction); the --dry-run and VFR-phone rows dropped (covered by the shared-flag line and a gotcha). Checked with a blind behaviour test, original vs trimmed, on 10 prompts: trimmed 60/60, original 56/60. test_delivery: the repair source is one TS appended to itself byte for byte, not bytes flipped inside packets; ffprobe 5.1.1 could not open the latter. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Repair test: fall back to MPEG-PS where ffprobe cannot read MPEG-TS The static ffprobe 5.1.1 build segfaults on every MPEG-TS file (the same reason test_contract gates its TS cut test), so the appended-recording source is written as MPEG-PS there; both containers are what a recorder appends to, and the assertions run unchanged on every runner. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Repair test: give the MPEG-PS fallback damage every ffmpeg reports On the static 5.1.1 build an appended program stream decoded without a timestamp warning, so nothing was counted before the repair. The PS copy now also has bytes flipped inside packs of its middle half; the MPEG-2 decoder reports those as decode errors on every version. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Docs: multicam energy switch is a loudness proxy; why 啊 is not a default filler; link ffmpeg-full PATH setup - README / gotchas: `--switch energy` picks the loudest camera, not a recognised speaker; it needs each speaker on their own camera's mic. - design-decisions / scripts.md: Chinese 啊 is out of the zh list for the same reason as `like` (a sentence-final particle far more often than a hesitation); `--filler-extra 啊` adds it. - README HDR section: the zscale note links to Install for putting the keg-only ffmpeg-full on PATH. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Drop color.py --tonemap bt2390 (never worked); bring README in line with the tree color.py: FFmpeg's CPU tonemap filter has no BT.2390 curve in any version (only tonemap_opencl and libplacebo do), so --tonemap bt2390 always failed. The choices are now the filter's own curves; a test checks every choice against the ffmpeg under test and that bt2390 is refused. README, checked against the code: - CI runs on pull requests and pushes to main, not every push; the FFmpeg list includes the Ubuntu 26.04 / 8.0.1 job. - templates include audiogram; --template all renders the seven video destinations, not every template. - doctor's fonts.scripts covers bn, ta, lo and latin too. - --crf was removed except on export.py; --codec/--quality are not on export.py, render.py or batch.py. - output_schema names output_identity (every tool but probe). tests: the Windows skip for the POSIX shim was on the wrong test since #269; it is back on test_dry_run_never_runs_ffmpeg_and_writes_nothing. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * check.py --compare: pair only frames both files have; RGB and rotated inputs; empty --expect-identity - ssim/psnr ran with eof_action=repeat, so a shorter output's last frame was measured against the rest of the reference (a mean of 0.90 for 52 identical frames, the worst frame timed past the file's end): shortest=1. - An RGB file printed "SSIM R:"/"PSNR r:", which the parser missed ("could not measure"), and the value depended on argument order: both inputs go through format=yuv444p and the parser takes any channel letter. - A clip stored sideways was compared by its stored size although ffmpeg auto-rotates: the size check uses the displayed size. - --expect-identity "" (an unset shell variable) skipped the identity row; it is now malformed (kind input), parsed before any measurement. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: measure stream spans from packets; VP9; --codec; dropped tracks. cut.py --remove outside the file - The A/V end check read stated stream lengths: Matroska/WebM state none, so a clip whose sound stopped 1.5 s early joined as verified; MPEG-TS states an estimate, so every AAC .ts was refused with wrong numbers. Each stream's span now comes from its first and last packet. - VP8/VP9 carry no extradata in WebM: identical clips were refused as "codec configuration differs" (and cut.py --remove/--segments fell back to an x264 re-encode into .webm, which ffmpeg refuses). - --copy with --codec, --quality or --hw ignored them silently; refused now. - A copy join drops subtitle/data tracks: dropped_non_av_streams says so. - cut.py --remove 15-20 on a 4 s file kept a range past the end and removed nothing; a range outside the file is refused. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * export.py --max-mb with gif / huge limits; --repair damage counting; render.py --stop-after keeps its output - --preset gif returned before --max-mb was applied, so a 0.7 MB GIF passed a 1 KB limit as verified: refused. A limit so large it constrains nothing is refused as input instead of failing inside libx264. - --repair counted "Packet corrupt (... dts = ...)" as a timestamp warning and "Last message repeated N times" as a decode error, so a remux that made timestamps worse could pass: repeat+warning prints each occurrence and corrupt-packet lines count as damage. - render.py --stop-after (before export) without --work removed the work directory at exit, deleting the file the result named; it is kept. - docs/contract.md lists size_over_limit and repair_incomplete. - CHANGELOG: the fixes found reviewing 2.24.0. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: refuse when a stream span cannot be measured; refuse any --quality - _stream_span skipped a packet without a time and measured the rest, which could invent (or hide) an A/V gap; it returns None, and join_copy refuses the copy when either span is unmeasured instead of skipping the clip. - --quality was refused only when it differed from the default, so --quality 18 passed silently; any --quality given is refused. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: read packet spans as JSON with a dts fallback; rotation fixture for FFmpeg 5.x The ffmpeg 5.1.1 job refused every copy join as "could not measure": a row of the csv packet listing did not parse as a time there. The spans are read from ffprobe's JSON (only packet entries), and a packet without a pts falls back to its dts; a packet with neither still makes the span unmeasured. -display_rotation is FFmpeg 6.0+; the rotated --compare fixture falls back to the rotate tag, and the test asserts the rotation was written. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: read a long clip's packet span at its head and tail only; test the dts fallback _stream_span listed every packet of the first video and audio stream of each clip to keep two numbers. A clip longer than 30 s is now read at its first 64 packets (past B-frame reordering) and its last 10 s; a test checks that this gives the same span as reading every packet. A packet with no pts is placed by its dts, now pinned by a test. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: seek a long clip's tail from its own start time -read_intervals takes timestamps, so the tail read started at duration - 10 s on the absolute clock. A clip whose clock starts late (an MPEG-TS hours in) read from its first packet instead, and one starting below -10 s read nothing and was refused. The seek is start_time + duration - 10 s; a test checks a clip starting at 100 s. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Restore CHANGELOG.md to main after the 2.24.1 release bump The merge kept the 2.24.1 Fixed entries under Unreleased as well as in the released section. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * 3.0.0: require FFmpeg 6.1+ and Python 3.10+; apply the cut.py defaults decided for 3.0 Work in progress (docs, the MCP snapshot and the tests that pin the 2.x cut defaults follow). - FFmpeg 5.x (deprecated in 2.14.0) and Python 3.9 (2.10.9) move from `deprecated` to `removed`; requirements are >=6.1 and >=3.10. Every tool refuses an ffmpeg/ffprobe it reads as older than 6.1 (kind missing_tool, reason ffmpeg_unsupported); doctor lists `unsupported` and reports ok false; the installer skips Python < 3.10 and prints the FFmpeg line. drawtext's 5.x/6.0 spellings and CI's 5.1.1 static job are gone; CI's Python floor is 3.10. - cut.py: edit list kept on a single .mp4/.mov copy (--no-edit-list for 2.x), SDR HEVC stays HEVC (--no-keep-hevc), --vfr-guard sampled; precision is the least exact segment's and keyframe_snapped equals start_snapped. contract_version 2.0. - package.json / plugin.json / roadmap 3.0.0 (a major is bumped by hand), CHANGELOG 3.0.0, the 2.14.0 release date and the 3.0.0 window waiver in the deprecation-window test. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * 3.0.0 docs: work in progress Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * 3.0.0 docs: work in progress Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * 3.0.0 docs: work in progress Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * v3.0: tests for the new cut defaults, sampled guard on seeked TS windows, structured ffmpeg_unsupported refusal Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * design-decisions: point stale test names at their 3.0 successors Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * probe: report the source's start timecode (video stream, tmcd, then format) Requested by nle-bridge-skill (roadmap step 12) for OTIO/EDL/FCP7 source timecodes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * test_asr: pick the engine call out of the fake's calls (the 6.1 floor reads ffprobe -version first) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Review: timecode minutes/seconds 00-59; ffprobe floor in doctor; git builds refused only below 6; README edit-list wording; sampled guard's keyframe start in the contract Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Review: doctor marks every tool unusable below the floor; an old ffprobe is named even when ffmpeg is old too Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * ci: spread the suite over three machines per platform (run_parallel.py --part I/N) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * run_parallel --part: bound N by the work items; test that every shard lands in exactly one part Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Failures say why: ffmpeg error reasons and hints, no orphan ffmpeg after SIGKILL, no blind re-encode, silent template renders, ASR missing_tool Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Review: one encoder_missing entry, asr_engine_missing under error.reason, render dry run reads sources and counts audio.replace Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: a talking-head reframe demo from AI-generated footage, on the landing page Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * silence --filler --transcribe: prime Whisper with the filler list; filler_en demo from AI-generated English speech Whisper leaves uh/um out of its text unless its context shows them, so --filler found none on clean speech. The filler list is passed as the initial prompt to whisper.cpp, faster-whisper and openai-whisper. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Pages: the two AI-footage demos as MP4 with sound, built with grid/graphics/join/look Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * demos: the gallery intro names the two AI-footage demos Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs/demos.md: keep render_project's --export-timeline line (a partial --docs rebuild dropped it) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Pages: the reframe label and aria-label give the source's real size; the lede names pages_videos.py Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: SKILL.md restructured, no behavior change SKILL.md keeps Quick start, Workflow 0-8, ask/assume, the does-not-decide line, the report template, the request table (one command per row) and the one-line gotchas. The long forms moved unchanged to references/flags.md, workflow.md, report.md, requests.md and timestamps.md, shipped via package.json files. Description cut to three sentences (924 -> 399 chars); trigger set 55/55 before and after. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: restore three SKILL.md rules the restructure cut; fix moved-text references - step 6 verifies from the writer's --json probe or probe.py (a routing check on agent_prompts_24 lost the verify step on ml1 without it) - step 7 allows --overwrite for a file the user asked to replace - the 60 s highlight row names cut.py --segments as the step that renders - references/flags.md points at SKILL.md's request table, not "the table below" - references/gotchas.md index links are same-file anchors - evals/results/routing-skill-md-restructure.json: old 93/99, new 94/99 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: step 5 keeps --expect-identity; highlight row and --from wording - step 5 runs check.py with --expect-identity S:M (the writer's output_identity), as before the restructure - the highlight row says picks.txt's lines join with commas for --segments - timestamps.md: only broll.py --from takes a time; render.py --from takes a stage name Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs --------- Co-authored-by: Claude <noreply@anthropic.com> | 18 days ago | |
docs: SKILL.md restructured, no behavior change (#392) * Add join.py --copy (stream-copy join of matching clips) and cut.py --remove join.py --copy --transition none joins clips by the concat demuxer with -c copy when their first video and audio streams match (cut.py's stream signature, now in _common.probe), no clip hides edit-list pre-roll, and each clip's picture and sound end within 50 ms. A mismatch is refused by name (join_copy_mismatch). The join is measured beside the output -- video packet count and length against the clips' -- before it is renamed into place. cut.py --remove A-B,... is the inverse of --segments: the ranges are dropped and the rest is cut exactly as the equivalent --segments would be. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Add output_identity, check.py --expect-identity and check.py --compare Every writing tool reports output_identity {size, mtime_ns} for the file it wrote; check.py --expect-identity SIZE:MTIME_NS FAILs an identity row when the file at the path is another one (an older file a failed run left, or a later run's), and check.py reports input_identity. SKILL.md's delivery check passes it. check.py --compare REF measures SSIM (mean and worst frame) and PSNR against a reference in one pass, WARN under a mean SSIM of 0.98. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Add export.py --max-mb: a two-pass x264 encode that fits a size, measured The preset's x264 line takes the video bitrate the running time leaves after the audio (97% of the budget) in place of its CRF and runs in two passes; the written size is a file_size verification step, with a size_over_limit warning when the encoder overshot. Refused for non-x264 presets and --hw, with --crf, without a duration, and under 100 kb/s for the picture. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Add export.py --preset copy --repair: remux a damaged file and count the damage Timestamps regenerated, corrupt packets dropped, demuxing past errors, negative timestamps shifted to zero; the input is only read. Both files are decoded once at -v warning and the timestamp warnings and other errors counted (repair: {before, after}); the repair step fails if either grew, and errors a remux cannot fix are a repair_incomplete warning. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * export.py --repair: judge the remux on timestamp warnings Decode errors are log lines whose count changes with the container (one full run logged 51 before and 53 after), so they are reported before and after but no longer fail the verification step; timestamp warnings, the part a remux rewrites, do. The test source is encoded single-threaded so its damage is the same every run. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Label the join copy, identity, compare, max-mb and repair additions 2.24 2.23.0 shipped with check.py's flashes and first-frame rows; these land in the next minor. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Address review: cover art, timeline times, copy-join guards, --remove edges, --max-mb - check.py --compare reads V:0 (not cover art), times the worst frame by its presentation time, counts frames from packets when nb_frames is missing, prints that note in plain text, runs under --dry-run when both files exist, and starts each main() with no leftover notes. - join.py --copy compares the first non-cover video stream, judges A/V alignment on each stream's end (start + length), includes colour matrix and range in the signature, refuses --sample-rate/--channels, fails the packet check when nothing was counted, and reports short and silent clips as the re-encoding join does. - cut.py --remove refuses any explicit --start and --snap beats, and clamps a range that starts before 0. - export.py --max-mb refuses nan and inf. - contract: join is conditional on audio; check notes name --compare. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Add a cut.py --remove demo; damage only video payload in the repair test The repair test overwrote bytes anywhere in the middle of the TS, which on ffmpeg 5.1.1 hit data its ffprobe could not open the file past; it now flips one byte in 40 video payload packets, never a TS header, the PAT/PMT or audio. The demo shows --remove 2-4 on the gallery's clip. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: refuse a different stream layout and a start offset; --compare pairs by index - The concat demuxer maps each clip's packets by stream index, so a clip whose streams are laid out differently (a cover picture between video and sound) is refused before the signatures are compared. - Picture and sound must start, as well as end, within 50 ms of each other in each clip. - check.py --compare re-times both pictures to their frame index before ssim/psnr, which otherwise pair frames by timestamp. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * check.py --expect-identity: grade an existing file under --dry-run too The identity was skipped for every dry run; it is skipped only for a pending placeholder now, as the content rows are. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy checks starts without stream lengths; --expect-identity reads twice - Matroska/WebM state no stream length, so the A/V guard compared nothing there; the starts are now compared whenever both streams exist and the ends only where both lengths are known. - check.py reads the input's identity before measuring and again after: a file replaced mid-check FAILs, and a path that holds no regular file is a FAIL row rather than no row. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * check.py: read the identity after --compare; note a mid-check change without --expect-identity The second identity read now follows every measurement, --compare's pass included, so a replacement during that pass is caught. Without --expect-identity no row is added (the row set is unchanged), but a file that changed while it was being checked is named in notes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Route the 2.24 features from SKILL.md; make the repair test open on every ffprobe SKILL.md: routing rows for export --max-mb, cut --remove, join --copy, export --preset copy --repair and check --compare; the report format moved above the routing table (inside the first ~5000 tokens kept after compaction); the --dry-run and VFR-phone rows dropped (covered by the shared-flag line and a gotcha). Checked with a blind behaviour test, original vs trimmed, on 10 prompts: trimmed 60/60, original 56/60. test_delivery: the repair source is one TS appended to itself byte for byte, not bytes flipped inside packets; ffprobe 5.1.1 could not open the latter. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Repair test: fall back to MPEG-PS where ffprobe cannot read MPEG-TS The static ffprobe 5.1.1 build segfaults on every MPEG-TS file (the same reason test_contract gates its TS cut test), so the appended-recording source is written as MPEG-PS there; both containers are what a recorder appends to, and the assertions run unchanged on every runner. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Repair test: give the MPEG-PS fallback damage every ffmpeg reports On the static 5.1.1 build an appended program stream decoded without a timestamp warning, so nothing was counted before the repair. The PS copy now also has bytes flipped inside packs of its middle half; the MPEG-2 decoder reports those as decode errors on every version. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Docs: multicam energy switch is a loudness proxy; why 啊 is not a default filler; link ffmpeg-full PATH setup - README / gotchas: `--switch energy` picks the loudest camera, not a recognised speaker; it needs each speaker on their own camera's mic. - design-decisions / scripts.md: Chinese 啊 is out of the zh list for the same reason as `like` (a sentence-final particle far more often than a hesitation); `--filler-extra 啊` adds it. - README HDR section: the zscale note links to Install for putting the keg-only ffmpeg-full on PATH. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Drop color.py --tonemap bt2390 (never worked); bring README in line with the tree color.py: FFmpeg's CPU tonemap filter has no BT.2390 curve in any version (only tonemap_opencl and libplacebo do), so --tonemap bt2390 always failed. The choices are now the filter's own curves; a test checks every choice against the ffmpeg under test and that bt2390 is refused. README, checked against the code: - CI runs on pull requests and pushes to main, not every push; the FFmpeg list includes the Ubuntu 26.04 / 8.0.1 job. - templates include audiogram; --template all renders the seven video destinations, not every template. - doctor's fonts.scripts covers bn, ta, lo and latin too. - --crf was removed except on export.py; --codec/--quality are not on export.py, render.py or batch.py. - output_schema names output_identity (every tool but probe). tests: the Windows skip for the POSIX shim was on the wrong test since #269; it is back on test_dry_run_never_runs_ffmpeg_and_writes_nothing. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * check.py --compare: pair only frames both files have; RGB and rotated inputs; empty --expect-identity - ssim/psnr ran with eof_action=repeat, so a shorter output's last frame was measured against the rest of the reference (a mean of 0.90 for 52 identical frames, the worst frame timed past the file's end): shortest=1. - An RGB file printed "SSIM R:"/"PSNR r:", which the parser missed ("could not measure"), and the value depended on argument order: both inputs go through format=yuv444p and the parser takes any channel letter. - A clip stored sideways was compared by its stored size although ffmpeg auto-rotates: the size check uses the displayed size. - --expect-identity "" (an unset shell variable) skipped the identity row; it is now malformed (kind input), parsed before any measurement. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: measure stream spans from packets; VP9; --codec; dropped tracks. cut.py --remove outside the file - The A/V end check read stated stream lengths: Matroska/WebM state none, so a clip whose sound stopped 1.5 s early joined as verified; MPEG-TS states an estimate, so every AAC .ts was refused with wrong numbers. Each stream's span now comes from its first and last packet. - VP8/VP9 carry no extradata in WebM: identical clips were refused as "codec configuration differs" (and cut.py --remove/--segments fell back to an x264 re-encode into .webm, which ffmpeg refuses). - --copy with --codec, --quality or --hw ignored them silently; refused now. - A copy join drops subtitle/data tracks: dropped_non_av_streams says so. - cut.py --remove 15-20 on a 4 s file kept a range past the end and removed nothing; a range outside the file is refused. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * export.py --max-mb with gif / huge limits; --repair damage counting; render.py --stop-after keeps its output - --preset gif returned before --max-mb was applied, so a 0.7 MB GIF passed a 1 KB limit as verified: refused. A limit so large it constrains nothing is refused as input instead of failing inside libx264. - --repair counted "Packet corrupt (... dts = ...)" as a timestamp warning and "Last message repeated N times" as a decode error, so a remux that made timestamps worse could pass: repeat+warning prints each occurrence and corrupt-packet lines count as damage. - render.py --stop-after (before export) without --work removed the work directory at exit, deleting the file the result named; it is kept. - docs/contract.md lists size_over_limit and repair_incomplete. - CHANGELOG: the fixes found reviewing 2.24.0. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: refuse when a stream span cannot be measured; refuse any --quality - _stream_span skipped a packet without a time and measured the rest, which could invent (or hide) an A/V gap; it returns None, and join_copy refuses the copy when either span is unmeasured instead of skipping the clip. - --quality was refused only when it differed from the default, so --quality 18 passed silently; any --quality given is refused. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: read packet spans as JSON with a dts fallback; rotation fixture for FFmpeg 5.x The ffmpeg 5.1.1 job refused every copy join as "could not measure": a row of the csv packet listing did not parse as a time there. The spans are read from ffprobe's JSON (only packet entries), and a packet without a pts falls back to its dts; a packet with neither still makes the span unmeasured. -display_rotation is FFmpeg 6.0+; the rotated --compare fixture falls back to the rotate tag, and the test asserts the rotation was written. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: read a long clip's packet span at its head and tail only; test the dts fallback _stream_span listed every packet of the first video and audio stream of each clip to keep two numbers. A clip longer than 30 s is now read at its first 64 packets (past B-frame reordering) and its last 10 s; a test checks that this gives the same span as reading every packet. A packet with no pts is placed by its dts, now pinned by a test. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * join.py --copy: seek a long clip's tail from its own start time -read_intervals takes timestamps, so the tail read started at duration - 10 s on the absolute clock. A clip whose clock starts late (an MPEG-TS hours in) read from its first packet instead, and one starting below -10 s read nothing and was refused. The seek is start_time + duration - 10 s; a test checks a clip starting at 100 s. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Restore CHANGELOG.md to main after the 2.24.1 release bump The merge kept the 2.24.1 Fixed entries under Unreleased as well as in the released section. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * 3.0.0: require FFmpeg 6.1+ and Python 3.10+; apply the cut.py defaults decided for 3.0 Work in progress (docs, the MCP snapshot and the tests that pin the 2.x cut defaults follow). - FFmpeg 5.x (deprecated in 2.14.0) and Python 3.9 (2.10.9) move from `deprecated` to `removed`; requirements are >=6.1 and >=3.10. Every tool refuses an ffmpeg/ffprobe it reads as older than 6.1 (kind missing_tool, reason ffmpeg_unsupported); doctor lists `unsupported` and reports ok false; the installer skips Python < 3.10 and prints the FFmpeg line. drawtext's 5.x/6.0 spellings and CI's 5.1.1 static job are gone; CI's Python floor is 3.10. - cut.py: edit list kept on a single .mp4/.mov copy (--no-edit-list for 2.x), SDR HEVC stays HEVC (--no-keep-hevc), --vfr-guard sampled; precision is the least exact segment's and keyframe_snapped equals start_snapped. contract_version 2.0. - package.json / plugin.json / roadmap 3.0.0 (a major is bumped by hand), CHANGELOG 3.0.0, the 2.14.0 release date and the 3.0.0 window waiver in the deprecation-window test. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * 3.0.0 docs: work in progress Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * 3.0.0 docs: work in progress Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * 3.0.0 docs: work in progress Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * v3.0: tests for the new cut defaults, sampled guard on seeked TS windows, structured ffmpeg_unsupported refusal Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * design-decisions: point stale test names at their 3.0 successors Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * probe: report the source's start timecode (video stream, tmcd, then format) Requested by nle-bridge-skill (roadmap step 12) for OTIO/EDL/FCP7 source timecodes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * test_asr: pick the engine call out of the fake's calls (the 6.1 floor reads ffprobe -version first) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Review: timecode minutes/seconds 00-59; ffprobe floor in doctor; git builds refused only below 6; README edit-list wording; sampled guard's keyframe start in the contract Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Review: doctor marks every tool unusable below the floor; an old ffprobe is named even when ffmpeg is old too Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * ci: spread the suite over three machines per platform (run_parallel.py --part I/N) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * run_parallel --part: bound N by the work items; test that every shard lands in exactly one part Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Failures say why: ffmpeg error reasons and hints, no orphan ffmpeg after SIGKILL, no blind re-encode, silent template renders, ASR missing_tool Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Review: one encoder_missing entry, asr_engine_missing under error.reason, render dry run reads sources and counts audio.replace Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: a talking-head reframe demo from AI-generated footage, on the landing page Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * silence --filler --transcribe: prime Whisper with the filler list; filler_en demo from AI-generated English speech Whisper leaves uh/um out of its text unless its context shows them, so --filler found none on clean speech. The filler list is passed as the initial prompt to whisper.cpp, faster-whisper and openai-whisper. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Pages: the two AI-footage demos as MP4 with sound, built with grid/graphics/join/look Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * demos: the gallery intro names the two AI-footage demos Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs/demos.md: keep render_project's --export-timeline line (a partial --docs rebuild dropped it) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Pages: the reframe label and aria-label give the source's real size; the lede names pages_videos.py Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: SKILL.md restructured, no behavior change SKILL.md keeps Quick start, Workflow 0-8, ask/assume, the does-not-decide line, the report template, the request table (one command per row) and the one-line gotchas. The long forms moved unchanged to references/flags.md, workflow.md, report.md, requests.md and timestamps.md, shipped via package.json files. Description cut to three sentences (924 -> 399 chars); trigger set 55/55 before and after. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: restore three SKILL.md rules the restructure cut; fix moved-text references - step 6 verifies from the writer's --json probe or probe.py (a routing check on agent_prompts_24 lost the verify step on ml1 without it) - step 7 allows --overwrite for a file the user asked to replace - the 60 s highlight row names cut.py --segments as the step that renders - references/flags.md points at SKILL.md's request table, not "the table below" - references/gotchas.md index links are same-file anchors - evals/results/routing-skill-md-restructure.json: old 93/99, new 94/99 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: step 5 keeps --expect-identity; highlight row and --from wording - step 5 runs check.py with --expect-identity S:M (the writer's output_identity), as before the restructure - the highlight row says picks.txt's lines join with commas for --segments - timestamps.md: only broll.py --from takes a time; render.py --from takes a stage name Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs --------- Co-authored-by: Claude <noreply@anthropic.com> | 18 days ago | |
chore: tests, CI, installer, evals and docs findings from the audit (#489) Audit findings in #467 and #466 that do not touch tool code: test isolation and sweep coverage, a release bump that is dropped (and releases nothing) when main moved, a Pages rebuild after each release, installer payload/argument/scratch-dir fixes, eval graders and demo commands, and docs corrected against the tree. | 6 days ago | |
0.8.1: skill craft — restructured SKILL.md, agent-run evals, fixes found in transcripts (#11) * SKILL.md restructure: trigger-focused description, workflow + decisions + report format; script reference and device notes moved to references/ - description states when to trigger, not just what exists - body 402 -> 169 lines; per-script CLI reference in references/scripts.md, real-device notes in references/devices.md (installed and packaged) - new sections: what to ask vs assume, report format, things that look right but are wrong - check.py: untagged 8-bit video passes the colour check (players assume BT.709); agents were adding a pointless --retag step - look.py: --at frames are stamped with the requested time, not the post-seek 0 Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * SKILL.md: do not normalise near-silent ambience to speech targets Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * 0.8.1: skill-craft release — agent-run evals, check step reworded, eval artifacts in evals/ Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs --------- Co-authored-by: Claude <noreply@anthropic.com> | 2 months ago | |
docs: eval 23b at 1.25.0 — karaoke-style word + faststart-fix coverage (#280) Continues eval 23 (closed partial at 1.19.3, evals/results/iteration-23-partial.json) with real-execution prompts for the two features shipped since: caption.py --karaoke-style word (#276) and the +faststart fixes (#275/#277/#279). - evals/agent_prompts_24.json: two new prompts, kw1 (karaoke-style word, phrased after the #276 reporter's own complaint that the highlight "just changed colour") and ff1 (a YouTube template delivery through the loudness-normalize path, probing for a faststart regression). - Both run for real (actual ffmpeg encodes, not simulated) and graded with the repo's own evals/grade_runs_24.py: routing 100%, report format 2/2, disclosure 2/2, real execution 2/2. kw1's generated ASS carries \fscx112\fscy112 on the active word (genuinely bigger, not just recoloured); ff1's output has moov before mdat after the loudnorm re-encode and check.py's loudness/true-peak rows PASS. No regression found. - docs/roadmap.md: corrects the stale "released version today" line (1.24.1 -> 1.25.0, which is only a chore: faststart-path bump, #279), adds table rows for 1.24.1/1.25.0, and a narrative section for eval 23b. - The Opus-focused-grader and 55-prompt trigger-judge passes are NOT run here: this sandbox has no independent/non-Claude model access, the same blocker iteration-23-partial already documented (see also docs/design-decisions.md's "The evaluation is Claude-only"). Documented as an open gap, not faked. Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs Co-authored-by: Claude <noreply@anthropic.com> | 1 month ago | |
fix: route "timeline for Premiere" from multicam through --write-project; add eval 25 kit (#293) * evals: iteration 25 (TTS joins, timeline handoff routing; trigger 3/3, agent runs not run) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix(docs): multicam --edl is a cut list; route editors through --write-project and --export-timeline Found reading eval 25's tl2 prompt: an agent asked for a Premiere timeline of a multicam edit would reach for multicam's --edl, which writes a START-END list no editor opens. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs --------- Co-authored-by: Claude <noreply@anthropic.com> | 1 month ago | |
Filler words in Korean, Chinese, Russian, Dutch, Polish and Turkish; Japanese fillers split across whisper tokens; Traditional Chinese fonts (#371) * silence.py --filler: Korean, Chinese, Russian, Dutch, Polish and Turkish lists Built-in hesitation lists for ko, zh, ru, nl, pl and tr, with words that are as often content (저, 那个, ну) named as discourse markers instead. Whisper splits Japanese and Chinese into tokens, not words, so a listed word may now span up to four tokens (えー + と is えーと), and in an unspaced script a match needs punctuation or a 0.1 s pause on each side (あの人 is kept). whisper.cpp no longer glues a whole CJK segment into one word. A two-word --filler-extra such as "you know" matches. Language names and ISO 639-2 codes in a --words document map to the right list. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * caption/graphics: a --lang naming Traditional Chinese resolves a Traditional face Han-only text always resolved a Simplified Chinese face, so a zh-TW or zh-HK caption was drawn in the mainland glyph forms. --lang zh-TW, zh-HK, zh-MO, zh-Hant or yue now looks the font up under zh-hant (Noto Sans CJK TC, Microsoft JhengHei); with no Traditional face installed the Simplified one draws it and font_variant_missing is recorded. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * evals: Korean, Chinese, Japanese and Taiwan prompts for the 2.18 multilingual work; SKILL.md names --lang zh-TW agent_prompts_26.json (not yet run): fk1 Korean fillers with 저 kept, fz1 Chinese fillers at a boundary with 金额 kept, fj2 Japanese えーと/えっと split across whisper tokens, tw1 a Taiwan caption that must pass --lang zh-TW. SKILL.md's CJK line names zh-TW so an agent can find it. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Filler runs stop at punctuation and reject reversed timings; Unifont no longer blocks the Simplified fallback; Korean filler demo - filler_spans: a run with punctuation between its tokens (えー、 + と) is not the listed えーと, so only えー is cut; a run whose last token ends before its first starts counts as untimed instead of emitting an inverted removal span. - script_font_for_text: a Traditional lookup that finds only Unifont now falls back to a real Simplified face (recorded as font_variant_missing). - references/scripts.md: the built-in lists table gains ko, zh, nl, pl, ru, tr, and a test keeps it equal to FILLER_WORDS. - demos: filler_korean (silence.py --filler on a Korean transcript). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * Filler runs end at a pause; whisper.cpp keeps 嗯,呃 apart; a .ttc names the family that ranked - filler_spans: a pause of FILLER_BOUNDARY_GAP inside an unspaced run is a word boundary, like punctuation, so えー + pause + と cuts only えー. - _unspaced_script covers Han extensions B-H (U+20000-U+323AF). - whisper.cpp: a punctuation token stays with the word before it and ends it, so 嗯 + , + 呃 are two words both lists can match, not 嗯,呃. - fonts: a file that lists several families (Noto CJK's JP/KR/SC/TC) returns the one that matched the preference, so the ASS Fontname names the TC face. - evals 26: --filler no longer matches --filler-extra; tw1 needs --lang zh-TW in command form. - The Unifont fallback test checks the returned file (CodeRabbit nitpick). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * A listed phrase that cannot be cut no longer hides the shorter filler inside it filler_spans tried the longest listed run first and, when that run was untimed or longer than max_word, counted it and skipped past it, so "um" under a held "um huh" was never tried. A rejected run now lets the shorter runs be tried; only when none of them can be cut is the longest one counted. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * A rejected filler run moves on by one token, so a listed word at its end is tried The previous change let the shorter runs at the same start be tried, but then skipped the whole rejected run, so "um" after a held "uhhh" in a listed "uhhh um" was never tried. The run is still counted once. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs --------- Co-authored-by: Claude <noreply@anthropic.com> | 20 days ago | |
Audio-only inputs: SKILL.md section, agent/trigger evals, grader checks, test (#16) Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs Co-authored-by: Claude <noreply@anthropic.com> | 2 months ago | |
FFmpeg 8 capability detection; audio extraction, audio join, sample-accurate trims, typed dynamics (0.9.1) (#20) doctor / contract: filter rows recognised by their io-spec token, so the two-character flag column of FFmpeg 8+ no longer hides every filter; unreadable listings yield `unknown`, never `missing`; fixtures captured from FFmpeg 6.1.1, 8.1.2 and 9.0.1. audio.py: audio extraction, --audio-stream, typed --compress/--limit/--gate. join.py: audio-only inputs joined as audio. cut.py: codec from the output extension, sample-accurate --accurate, precision reported. CI on pull_request with listing artifacts. Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs | 2 months ago | |
fix(evals): f05-unsupported taught the wrong answer -- reverse.py exists (#83) evals/agent_prompts_exec.json's f05-unsupported claimed "no reverse tool in the skill" and scored refusing a reverse request as correct -- scripts/reverse.py has existed the whole time and SKILL.md's own Request->script table routes "reverse this clip" to it. The eval also endorsed "a hand-written ffmpeg reverse... if run and verified" as an acceptable fallback, directly contradicting SKILL.md's and agent_prompts_24.json's explicit "never fall back to a raw ffmpeg invocation, even when asked directly" policy. - Replaced f05-unsupported with a genuinely unsupported case (a custom -filter_complex effects chain no script exposes), reusing the same scenario agent_prompts_24.json already uses consistently, so the two eval files no longer disagree with each other or with SKILL.md. - Added x06-reverse as a normal success case for the real reverse.py path, verified against a real fixture (reverse.py exits 0, JSON reports a valid probed output). - Left evals/results/exec-1.json's historical record unedited (it's a record of a run, not a living spec) but added a "superseded" note explaining f05's old grading was wrong and pointing at this fix, rather than silently rewriting a past result. Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs Co-authored-by: Claude <noreply@anthropic.com> | 2 months ago | |
chore: tests, CI, installer, evals and docs findings from the audit (#489) Audit findings in #467 and #466 that do not touch tool code: test isolation and sweep coverage, a release bump that is dropped (and releases nothing) when main moved, a Pages rebuild after each release, installer payload/argument/scratch-dir fixes, eval graders and demo commands, and docs corrected against the tree. | 6 days ago | |
docs: eval 20 at 1.17.2 (20 caption runs, the picture fixed) (#241) * evals: eval-20 trigger results and grader fixes (cs1 size regex, cs2 Spanish disclosure, cs3 rewrite is not honest) Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: eval 20 at 1.17.2 (20 caption runs, the picture fixed, 0/20 one-word-per-line) Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs --------- Co-authored-by: Claude <noreply@anthropic.com> | 2 months ago | |
chore: tests, CI, installer, evals and docs findings from the audit (#489) Audit findings in #467 and #466 that do not touch tool code: test isolation and sweep coverage, a release bump that is dropped (and releases nothing) when main moved, a Pages rebuild after each release, installer payload/argument/scratch-dir fixes, eval graders and demo commands, and docs corrected against the tree. | 6 days ago | |
chore: tests, CI, installer, evals and docs findings from the audit (#489) Audit findings in #467 and #466 that do not touch tool code: test isolation and sweep coverage, a release bump that is dropped (and releases nothing) when main moved, a Pages rebuild after each release, installer payload/argument/scratch-dir fixes, eval graders and demo commands, and docs corrected against the tree. | 6 days ago | |
fix: audit #463, #464 and #467 open items across media tools, delivery and packaging (#521) * fix: frame size, colour tags, delivery limits and verify's outputs (#463, #464) The items of #463 and #464 still real on 3.8.8, each reproduced on FFmpeg 6.1.1, swept for its class, fixed at the shared cause and pinned by a test that measures the written file: - #463 A153: an untagged BT.2020 transfer was re-encoded as HLG (all 10-bit lines); now bt2020-10 (decision.source_transfer). - #463 A161: --to-sdr --force named every non-HDR input "untagged input as PQ"; sibling: a fully untagged input failed kind ffmpeg (zscale, 6.1). - #463 A154: straighten --fit crop wrote 2 px short on each axis. - #463 A157: crop rounded an odd --x/--y to the chroma grid; sibling: redact left the asked rectangle's last column/row unredacted. - #463 A158: fit --method trim from 0 added -ss 0 and dropped subtitles. - #463 A159: look --compare / the sheet used an extensionless -o verbatim; sibling (every media tool): an extensionless output is refused as input (output_extension_missing) before ffmpeg, dry runs included. - #463 A162: proxy upscaled small sources; tiny --scale became full size. - #463 A163: sphere's default v-fov now keeps pixels square. - #463 A156: stabilize sends optzoom=0, so --zoom is the only zoom. - #464 A115: youtube presets conform a source above fps_max to rate/n. - #464 A112: --max-mb --normalize re-encodes sound at the preset's bitrate. - #464 A123: probe --field prints JSON, not Python repr. - #464 A121, A119, A120: verify stems never collide, --overwrite/--fast reach its steps, --keep alone keeps the outputs and names out_dir. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix: the open items of audit #467 (memory, distribution, contract size, gallery, eval runner) - A217: decode_pcm_mono returns array('f') (about 8 bytes a sample at the peak, not ~81) - A218: verify.py walks _common's MEDIA_EXT less the stills, as batch.py does - A220: contract --json --tool NAME narrows `tools` (55-85 KB instead of ~950 KB) - A222/A27/A36: negated `files` entries keep __pycache__/*.pyc out of the npm tarball; release_check.sh fails a tarball that has one - A224: 39 unused imports, two dead variables and two uncalled functions removed; an ast test fails on an unused import in the shipped modules - A243/A244: the cropdetect_crop and scenes_highlights demos run the measured value and stop when it is not the one docs/demos.md prints - A271/A274: run_agents_25.sh installs the shipped skill with bin/install.js --dir and exits 1 when the checkout's git status changed during a run A221 was already fixed (#466 A195); A219 is decided in docs/design-decisions.md. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * docs: changelog entries of #463/#464/#467 sit under Unreleased after the rebase onto 3.8.13 Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix: redact blurs a region four pixels wide on FFmpeg 7.1; the dry-run alpha test patches the decode CI of #521: boxblur on FFmpeg 7.1 refuses a chroma radius of 1 on a 2 px chroma plane (6.1 took it), so --mode blur on a 4 px wide region failed. And test_accel patched export._run_analysis, an unused alias this PR removed: the check never guarded anything; it patches alpha_stats, the decode a dry run must not run. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix: review findings on verify's outputs, probe --field and the eval runner - verify.py plans every file before any step runs and checks every name it will write (step outputs, caption's .srt/.ass, the cue file) against every collected input with refuse_output_is_input (a link or another case of the name included): FOLDER --out FOLDER --overwrite with clip.mp4 and clip_cut.mp4 replaced clip_cut.mp4. Refused as kind input, reason output_is_input, before anything is written. batch.py was swept: it already refuses this (#428), and delivers by rename. - apply_common sets args.json to the effective mode, so --json-brief alone prints the JSON document in probe --field, check, cropdetect, silence --list, report and verify (the sweep of the probe finding). - redact.py: the chroma radius is at least 1 wherever the chroma plane can hold it; --blur-strength 1 left colour-only detail unblurred. - evals/run_agents_25.sh: the checkout's state adds the tracked diff against HEAD and the untracked files' hashes; a RUN_DIR inside the checkout that git does not ignore is refused (exit 2) before anything is staged. - CHANGELOG and design-decisions: probe --field's missing field is an empty line without --json/--json-brief, null with either. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix: the agent eval's checkout state carries HEAD, and a snapshot that fails stops the run Review of #521: a run that committed its edit left git status clean and the diff against the new HEAD empty; and two failed snapshots compared equal. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * test: the committed-edit case of the agent eval commits a file that was clean Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix: a failed listing of untracked files fails the agent eval's snapshot Review of #521: ls-files sat in front of a pipe, so its failure was the pipeline's success. The list goes through a file and both statuses count. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs * fix: the agent eval's snapshot carries the index and hashes an untracked link without following it Review of #521: a run that only restaged a file left the status line and the diff against HEAD as they were; a dangling untracked link failed the snapshot (git hash-object opens the target). Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs --------- Co-authored-by: Claude <noreply@anthropic.com> | 4 days ago | |
Audio-only inputs: SKILL.md section, agent/trigger evals, grader checks, test (#16) Claude-Session: https://claude.ai/code/session_013tPrsGXLuaGU7t643QNUZs Co-authored-by: Claude <noreply@anthropic.com> | 2 months ago | |
chore: tests, CI, installer, evals and docs findings from the audit (#489) Audit findings in #467 and #466 that do not touch tool code: test isolation and sweep coverage, a release bump that is dropped (and releases nothing) when main moved, a Pages rebuild after each release, installer payload/argument/scratch-dir fixes, eval graders and demo commands, and docs corrected against the tree. | 6 days ago |
| Files | Last commit | Last update |
|---|---|---|
| 2 months ago | ||
| 18 days ago | ||
| 18 days ago | ||
| 6 days ago | ||
| 2 months ago | ||
| 1 month ago | ||
| 1 month ago | ||
| 20 days ago | ||
| 2 months ago | ||
| 2 months ago | ||
| 2 months ago | ||
| 6 days ago | ||
| 2 months ago | ||
| 6 days ago | ||
| 6 days ago | ||
| 4 days ago | ||
| 2 months ago | ||
| 6 days ago |