[vanceingalls]

What

A piped, non-quiet CLI render's @hf-result line now carries the render's motion blur summary, so a host that runs the CLI as a subprocess can read it. HyperFrames Desktop is that host: it sees only the exit code, @hf-progress lines and stderr, so it cannot report motion_blur_failed_frames for the blur default-on rollout gate.

@hf-result {"version":1,"operation":"render","reference_background":{...},"motion_blur":{"tier":"filter","movers":77,"failed_frames":0,"fingerprint":{...14 counts...}}}

Content-free

motion_blur is built by renderMotionBlurResult in telemetry/renderMotionBlur.ts: tier narrowed to capture/filter, and non-negative integers for movers, failed_frames and the 14 fingerprint counts (snake_case, the names the telemetry payload already uses). No reasonCode, path, selector or composition text can reach it; a test sets those on the summary and asserts the line does not contain them.

Behaviour

• Only the piped, non-quiet, successful local render prints the line, as before. --docker still prints none.
• The line is absent of motion_blur when the render has no blur summary, so a film without blur is byte-identical to today.
• The early return now fires only when there is neither a measured background nor a blur summary; the local render path always has both.
• Version stays 1: the README already tells consumers to ignore unknown fields, and I found no @hf-result consumer in this repo or in hyperframes-internal. README documents the new field.
• The fingerprint carries zero counts too; telemetry drops them. Easy to drop here if a consumer wants parity.

Tests

renderResult.test.ts (present for a blur render, absent without a summary, no free text, existing background cases unchanged) and execute.test.ts (the summary reaches the line from a local render): 30 tests across the three relevant files pass. Test reachability, comment ratchet, comment citations and tsc --noEmit in packages/cli exit 0.

Next

A release, then Desktop parses the line and bumps its pin (separate PR in hyperframes-internal).

🤖 Generated with Claude Code

[github-actions]

Edit accuracy: accurate 267 (base branch 267), smooth 94 of those

The gate passes.
Smoothness is reported in the artifact, not gated. A case fails only if it fails 2 of 3 runs.

Quarantined, measured but not gated (8)

• seqplainkeys-none-center-r0-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-center-r0-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-center-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-pct-r0-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-pct-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-px-r0-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-px-r30-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-px-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run

Unstable (2)

• seqplainkeys-none-px-r0-root-z100: tracking 0.06, pressJump 0, drop 0, reload 0, render 0.03, renderKey -, undo true, teleport true / tracking 0.06, pressJump 0, drop 0, reload 0, render 0.03, renderKey -, undo true, teleport true / tracking 0.06, pressJump 0, drop 0, reload 0, render 0.03, renderKey -, undo true, teleport true
• seqrepeat-none-center-r0-root-z100: tracking 0.05, pressJump 0, drop 0, reload 0, render 0.02, renderKey -, undo true, teleport true / tracking 0.05, pressJump 0, drop 0, reload 0, render 0.02, renderKey -, undo true, teleport true / tracking 0.05, pressJump 0, drop 0, reload 0, render 0.02, renderKey -, undo true, teleport true

[heygengenesis]

Verdict: APPROVE · 0 blocker · 0 important · 0 nits · head 31d1eb7

The content-free guarantee holds. renderMotionBlurResult builds every key from constants (MOTION_BLUR_FINGERPRINT_FIELDS through snake), never from the page object. tier passes an explicit two-value check. Every value goes through count(), and reasonCode/readEvents are never read. Docker returns { renderTimeMs }, so no line is printed, and batch rows are unchanged.

Scope

Traced the page values from readPageMotionBlur → summarize → aggregateMotionBlur → executeLocalRender → renderMachineResult. The values are integers, but the page supplies them and nothing bounds them: a composition that overrides window.__hfMotionBlur can choose any integer up to 2^53 for each field. Telemetry already carries those same values, so this PR doesn't change exposure. The motion-blur-only branch (no reference_background) can't be reached in production because referenceBackgroundResult always returns the field. CI is green; CodeQL and Windows render verification were still running when I posted.

— Genesis