feat: add modular content creation primitive endpoints#390

Open

recoupableorg wants to merge 32 commits intomainfrom

feature/content-primitive-endpoints

Commits on Apr 2, 2026

feat: add modular content creation primitive endpoints
sidneyswift
committed
fix: inline route segment config (Next.js 16 requires static analysis)
sidneyswift
committed
fix: address CodeRabbit review comments
sidneyswift
committed
fix: resolve all lint errors in new primitive files
sidneyswift
committed
refactor: make primitives run inline instead of triggering tasks
sidneyswift
committed
fix: enforce validateAuthContext at handler level in content primitives
sidneyswift
committed
feat: add POST /api/content/create/analyze (Twelve Labs video analysis)
sidneyswift
committed
fix: rename content/create/analyze to content/analyze
sidneyswift
committed
refactor: rename content primitive routes to verb-qualifier pattern
sidneyswift
committed
refactor: make content primitives generic + replace render with edit
sidneyswift
committed
fix: address CodeRabbit review — split text handler, DRY route factory
sidneyswift
committed
fix: make image_url optional in generate-video, add prompt field
sidneyswift
committed
chore: redeploy with FAL_KEY
sidneyswift
committed
chore: redeploy with updated FAL_KEY
sidneyswift
committed
fix: upgrade to nano-banana-2, auto-select t2i vs edit model
sidneyswift
committed
feat: add num_images, aspect_ratio, resolution to generate-image
sidneyswift
committed
feat: set optimal internal defaults for image generation
sidneyswift
committed
feat: auto-select Veo 3.1 model variant based on inputs
sidneyswift
committed
feat: full generate-video upgrade — extend mode, duration, resolution, audio
sidneyswift
committed
feat: expose missing params for transcribe + upscale, fix generate_audio default
sidneyswift
committed
feat: add mode param to generate-video with 6 modes
sidneyswift
committed
fix: correct text-to-video model ID (fal-ai/veo3.1, not /text-to-video)
sidneyswift
committed
fix: correct fal field mappings for reference and first-last modes
sidneyswift
committed
refactor: simplify endpoint paths per code review
sidneyswift
committed
chore: redeploy with updated RECOUP_API_KEY
sidneyswift
committed
chore: redeploy with new RECOUP_API_KEY for caption
sidneyswift
committed
fix: always send prompt field to fal (LTX lipsync requires it even when empty)
sidneyswift
committed
refactor: caption handler calls AI SDK directly instead of HTTP self-call
sidneyswift
committed
refactor: extract configureFal and buildFalInput (DRY + SRP)
sidneyswift
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: add modular content creation primitive endpoints#390

feat: add modular content creation primitive endpoints#390
recoupableorg wants to merge 32 commits intomainfrom
feature/content-primitive-endpoints

Commits on Apr 2, 2026

feat: add modular content creation primitive endpoints

fix: inline route segment config (Next.js 16 requires static analysis)

fix: address CodeRabbit review comments

fix: resolve all lint errors in new primitive files

refactor: make primitives run inline instead of triggering tasks

fix: enforce validateAuthContext at handler level in content primitives

feat: add POST /api/content/create/analyze (Twelve Labs video analysis)

fix: rename content/create/analyze to content/analyze

refactor: rename content primitive routes to verb-qualifier pattern

refactor: make content primitives generic + replace render with edit

fix: address CodeRabbit review — split text handler, DRY route factory

fix: make image_url optional in generate-video, add prompt field

chore: redeploy with FAL_KEY

chore: redeploy with updated FAL_KEY

fix: upgrade to nano-banana-2, auto-select t2i vs edit model

feat: add num_images, aspect_ratio, resolution to generate-image

feat: set optimal internal defaults for image generation

feat: auto-select Veo 3.1 model variant based on inputs

feat: full generate-video upgrade — extend mode, duration, resolution, audio

feat: expose missing params for transcribe + upscale, fix generate_audio default

feat: add mode param to generate-video with 6 modes

fix: correct text-to-video model ID (fal-ai/veo3.1, not /text-to-video)

fix: correct fal field mappings for reference and first-last modes

refactor: simplify endpoint paths per code review

chore: redeploy with updated RECOUP_API_KEY

chore: redeploy with new RECOUP_API_KEY for caption

fix: always send prompt field to fal (LTX lipsync requires it even when empty)

refactor: caption handler calls AI SDK directly instead of HTTP self-call

refactor: extract configureFal and buildFalInput (DRY + SRP)

Commits on Apr 3, 2026

fix: remove unused stream from analyze schema, DRY video OPTIONS handler

feat: add template support to all content primitives

feat: content V2 — edit route, template detail, malleable mode, MCP tools