chore: add designer agent + agents README index#31
Merged
Conversation
Add a `designer` subagent that owns the output-experience axis no existing agent covers — video pacing, callout/cursor/zoom placement, caption + .srt legibility, GIF exports, and CLI ergonomics — judged by watching a real render rather than reading the filtergraph. Add an agents/README.md dispatch index covering all five agents (when to reach for each, what each produces), framing the four reviewers as distinct axes on the same render (correct / proven / watches-well / safe-to-ship), their handoffs, and why there is deliberately no security-reviewer. Adapted from the pilot-forge agent setup; persona-engineer, reporter, and security-reviewer were intentionally not ported. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Ports two ideas from the pilot-forge agent setup, adapted to tutorial-forge:
New
designersubagent (.claude/agents/designer.md) — owns the one axis none of the existing agents cover: does the render watch well? The code-reviewer and qa-engineer prove the render is correct (drift, flash detect/trim, srt-vs-manifest cue count); the designer judges whether it's good to watch — video pacing, callout/cursor/zoom placement, caption +.srtlegibility, GIF exports, and CLI ergonomics (doctor, progress,StepError). It works by rendering something real (pnpm e2e/ the example-app harness) and inspecting frames/timing withffprobe/sips/ffmpeg, not by reading the filtergraph. Genuine bugs get handed back to code-reviewer/qa-engineer so it stays in the taste lane.New
agents/README.mdindex — a dispatch table for all five agents (when to reach for each, what each produces), framing the four reviewers as distinct axes on the same render (correct / proven / watches-well / safe-to-ship), documenting their handoffs, and recording why there is deliberately nosecurity-reviewer(TF has no runtime attack surface; the release-reviewer owns the real pre-ship risk).Why
The existing four agents had no index telling you when to reach for which or how they compose, and nothing owned the output-watching-experience axis — which is the whole product for a tutorial-video renderer.
Scope
Net-new files only; no existing agent or source touched.
Not ported (intentional)
persona-engineerandreporterare pilot-forge-specific;security-reviewerwas already deliberately replaced byrelease-reviewerin this repo.software-architectwas skipped — TF is past the greenfield phase where it earns its place.🤖 Generated with Claude Code