feat: handle ffmpeg-edit output in pollContentRuns#139
Open
recoup-coding-agent wants to merge 1 commit intomainfrom
Open
feat: handle ffmpeg-edit output in pollContentRuns#139recoup-coding-agent wants to merge 1 commit intomainfrom
recoup-coding-agent wants to merge 1 commit intomainfrom
Conversation
The ffmpeg-edit task returns `url` instead of `videoSourceUrl`. Add fallback to check `output.url` so the polling task can handle both create-content and ffmpeg-edit run outputs. Co-Authored-By: Paperclip <noreply@paperclip.ing>
📝 WalkthroughWalkthroughThe Changes
Estimated code review effort🎯 2 (Simple) | ⏱️ ~8 minutes Poem
🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
5 tasks
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/content/pollContentRuns.ts`:
- Around line 42-47: The code reads run.output into the local const output via a
type assertion but never validates it; add a Zod schema (e.g., OutputSchema)
that defines videoSourceUrl?: string, url?: string, captionText?: string and
uses refine() to require that at least one of videoSourceUrl or url is present,
then parse/validate run.output with OutputSchema.parse or safeParse before
treating the run as completed; if validation fails, do not return status:
"completed" (handle as still pending/errored) and only set videoUrl from the
validated output (prefer videoSourceUrl or fallback to url) when the schema
passes. Ensure these changes touch the same identifiers: the const output
assignment, any logic that returns status: "completed", and the code that reads
videoUrl/captionText.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 2ccca4a3-217b-4d0a-a2bc-6f145fe7202b
📒 Files selected for processing (1)
src/content/pollContentRuns.ts
Comment on lines
42
to
47
| const output = run.output as { | ||
| videoSourceUrl?: string; | ||
| captionText?: string; | ||
| /** ffmpeg-edit tasks return `url` instead of `videoSourceUrl`. */ | ||
| url?: string; | ||
| } | null; |
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
# Verify this file is currently using a type assertion for run.output
# and does not validate with Zod.
rg -n --type=ts 'run\.output\s+as\s+\{' src/content/pollContentRuns.ts
rg -n --type=ts 'from ["'"'"']zod["'"'"']|z\.object\(|safeParse\(' src/content/pollContentRuns.tsRepository: recoupable/tasks
Length of output: 99
🏁 Script executed:
cat -n src/content/pollContentRuns.tsRepository: recoupable/tasks
Length of output: 2253
🏁 Script executed:
cat -n src/content/sendContentCallback.tsRepository: recoupable/tasks
Length of output: 1792
🏁 Script executed:
cat -n src/tasks/pollContentRunTask.tsRepository: recoupable/tasks
Length of output: 1374
🏁 Script executed:
# Check how run.output is typed in the SDK
rg -n "runs\.poll\(" src/ -A 5 | head -20Repository: recoupable/tasks
Length of output: 1241
🏁 Script executed:
# Check if Zod is available in the project dependencies
cat package.json | grep -A 5 -B 5 "zod"Repository: recoupable/tasks
Length of output: 477
🏁 Script executed:
# Check for other schema definitions using Zod in the project
find src -name "*.ts" -type f | xargs grep -l "z.object\|z.string" | head -10Repository: recoupable/tasks
Length of output: 397
🏁 Script executed:
# Check an example Zod schema in the project
head -50 src/recoup/fetchTask.tsRepository: recoupable/tasks
Length of output: 1417
🏁 Script executed:
# Check if there are any tests for pollContentRuns to understand expected behavior
find . -name "*.test.ts" -o -name "*.spec.ts" | xargs grep -l "pollContentRuns" 2>/dev/nullRepository: recoupable/tasks
Length of output: 212
🏁 Script executed:
cat -n src/content/__tests__/pollContentRuns.test.tsRepository: recoupable/tasks
Length of output: 4345
🏁 Script executed:
# Verify the learnings about result.ok check for triggerAndWait
rg -n "triggerAndWait\|runs\.poll" src/content/pollContentRuns.ts -B 2 -A 2Repository: recoupable/tasks
Length of output: 42
Validate run output with Zod schema and require a URL before returning completed.
Lines 42–47 and 53 use a type assertion on run.output without schema validation. When both videoSourceUrl and url are absent or null, the code still returns status: "completed" with videoUrl: undefined, which can misrepresent success to external callbacks.
Per coding guidelines, "Use Zod for schema validation" for TypeScript files. Add a Zod schema with a refine() check to ensure at least one video URL field is present before marking the run as completed.
Proposed fix
import { runs } from "@trigger.dev/sdk/v3";
+import { z } from "zod";
import { logStep } from "../sandboxes/logStep";
const POLL_INTERVAL_MS = 30_000;
+const ContentRunOutputSchema = z
+ .object({
+ videoSourceUrl: z.string().url().optional(),
+ /** ffmpeg-edit tasks return `url` instead of `videoSourceUrl`. */
+ url: z.string().url().optional(),
+ captionText: z.string().optional(),
+ })
+ .refine(output => Boolean(output.videoSourceUrl ?? output.url), {
+ message: "Missing video URL in run output",
+ });
export type ContentRunResult = {
runId: string;
status: "completed" | "failed" | "timeout";
videoUrl?: string;
captionText?: string;
error?: string;
};
/**
* Waits for all Trigger.dev create-content runs to reach a terminal state
* using the native runs.poll() function, then maps results.
*/
export async function pollContentRuns(
runIds: string[],
): Promise<ContentRunResult[]> {
const settled = await Promise.allSettled(
runIds.map(runId =>
runs.poll(runId, { pollIntervalMs: POLL_INTERVAL_MS }),
),
);
return settled.map((result, i) => {
const runId = runIds[i];
if (result.status === "rejected") {
logStep(`Run poll failed: ${runId}`, false, { runId, error: result.reason });
return {
runId,
status: "failed" as const,
error: result.reason?.message ?? "Unknown error",
};
}
const run = result.value;
if (run.status === "COMPLETED") {
- const output = run.output as {
- videoSourceUrl?: string;
- captionText?: string;
- /** ffmpeg-edit tasks return `url` instead of `videoSourceUrl`. */
- url?: string;
- } | null;
+ const parsedOutput = ContentRunOutputSchema.safeParse(run.output ?? {});
+ if (!parsedOutput.success) {
+ logStep(`Run completed with invalid output: ${runId}`, false, {
+ runId,
+ issues: parsedOutput.error.issues,
+ });
+ return {
+ runId,
+ status: "failed" as const,
+ error: "Completed run missing valid output URL",
+ };
+ }
+ const output = parsedOutput.data;
logStep(`Run completed: ${runId}`, false, { runId });
return {
runId,
status: "completed" as const,
- videoUrl: output?.videoSourceUrl ?? output?.url,
- captionText: output?.captionText,
+ videoUrl: output.videoSourceUrl ?? output.url,
+ captionText: output.captionText,
};
}📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| const output = run.output as { | |
| videoSourceUrl?: string; | |
| captionText?: string; | |
| /** ffmpeg-edit tasks return `url` instead of `videoSourceUrl`. */ | |
| url?: string; | |
| } | null; | |
| import { runs } from "@trigger.dev/sdk/v3"; | |
| import { z } from "zod"; | |
| import { logStep } from "../sandboxes/logStep"; | |
| const POLL_INTERVAL_MS = 30_000; | |
| const ContentRunOutputSchema = z | |
| .object({ | |
| videoSourceUrl: z.string().url().optional(), | |
| /** ffmpeg-edit tasks return `url` instead of `videoSourceUrl`. */ | |
| url: z.string().url().optional(), | |
| captionText: z.string().optional(), | |
| }) | |
| .refine(output => Boolean(output.videoSourceUrl ?? output.url), { | |
| message: "Missing video URL in run output", | |
| }); | |
| export type ContentRunResult = { | |
| runId: string; | |
| status: "completed" | "failed" | "timeout"; | |
| videoUrl?: string; | |
| captionText?: string; | |
| error?: string; | |
| }; | |
| /** | |
| * Waits for all Trigger.dev create-content runs to reach a terminal state | |
| * using the native runs.poll() function, then maps results. | |
| */ | |
| export async function pollContentRuns( | |
| runIds: string[], | |
| ): Promise<ContentRunResult[]> { | |
| const settled = await Promise.allSettled( | |
| runIds.map(runId => | |
| runs.poll(runId, { pollIntervalMs: POLL_INTERVAL_MS }), | |
| ), | |
| ); | |
| return settled.map((result, i) => { | |
| const runId = runIds[i]; | |
| if (result.status === "rejected") { | |
| logStep(`Run poll failed: ${runId}`, false, { runId, error: result.reason }); | |
| return { | |
| runId, | |
| status: "failed" as const, | |
| error: result.reason?.message ?? "Unknown error", | |
| }; | |
| } | |
| const run = result.value; | |
| if (run.status === "COMPLETED") { | |
| const parsedOutput = ContentRunOutputSchema.safeParse(run.output ?? {}); | |
| if (!parsedOutput.success) { | |
| logStep(`Run completed with invalid output: ${runId}`, false, { | |
| runId, | |
| issues: parsedOutput.error.issues, | |
| }); | |
| return { | |
| runId, | |
| status: "failed" as const, | |
| error: "Completed run missing valid output URL", | |
| }; | |
| } | |
| const output = parsedOutput.data; | |
| logStep(`Run completed: ${runId}`, false, { runId }); | |
| return { | |
| runId, | |
| status: "completed" as const, | |
| videoUrl: output.videoSourceUrl ?? output.url, | |
| captionText: output.captionText, | |
| }; | |
| } | |
| }); | |
| } |
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/content/pollContentRuns.ts` around lines 42 - 47, The code reads
run.output into the local const output via a type assertion but never validates
it; add a Zod schema (e.g., OutputSchema) that defines videoSourceUrl?: string,
url?: string, captionText?: string and uses refine() to require that at least
one of videoSourceUrl or url is present, then parse/validate run.output with
OutputSchema.parse or safeParse before treating the run as completed; if
validation fails, do not return status: "completed" (handle as still
pending/errored) and only set videoUrl from the validated output (prefer
videoSourceUrl or fallback to url) when the schema passes. Ensure these changes
touch the same identifiers: the const output assignment, any logic that returns
status: "completed", and the code that reads videoUrl/captionText.
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
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
1 participant
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.
Summary
pollContentRunsto checkoutput.url(from ffmpeg-edit tasks) in addition tooutput.videoSourceUrl(from create-content tasks)Context
Part of REC-68: when users reply in existing content threads, the bot now triggers ffmpeg-edit tasks. The polling task needs to extract the video URL from the edit task output which uses a different field name.
Test plan
videoSourceUrl)urlfallback)🤖 Generated with Claude Code
Summary by cubic
Update
pollContentRunsto fall back tooutput.url(fromffmpeg-edit) whenoutput.videoSourceUrlis missing, so polling works for both edit and create-content runs. This fulfills REC-68 by supporting replies that triggerffmpeg-edittasks.Written for commit d4a6fe9. Summary will update on new commits.
Summary by CodeRabbit
Release Notes