Skip to content

Alex/refine ad product catalog validation - #48

Open
amurph491 wants to merge 12 commits into
mainfrom
alex/refine-ad-product-catalog-validation
Open

Alex/refine ad product catalog validation#48
amurph491 wants to merge 12 commits into
mainfrom
alex/refine-ad-product-catalog-validation

Conversation

@amurph491

Copy link
Copy Markdown
Collaborator

Description

Checklist

  • Tested against the Spotify Ads API with Codex, Claude Code, or Antigravity CLI
  • Existing skills still work as expected
  • SKILL.md frontmatter is valid (name, description, allowed-tools)
  • README or CHANGELOG updated (if user-facing change)
  • Curl commands follow API conventions documented in AGENTS.md

MiniRodz and others added 11 commits July 24, 2026 17:25
Adds a mandatory GET /ad_product_catalog step before entity creation
across all skills to validate field values against ad product rules.
When ad_product is UNSET/UNKNOWN/unspecified, AUCTION rules apply.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The catalog fetch happens in an early step, but agents lose track of it
by the time they reach the individual create curls for campaigns, ad
sets, and ads. This adds a validation reminder directly before each
create curl in drafts (4a/4b/4c), build-campaign (4a/4b/4c), and
clone (6a/6b/6c).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Three changes to the ad product catalog validation:

1. Always fetch the parent campaign to determine its ad_product before
   validating child entities (ad sets, ads). For multi-step flows,
   use the ad_product from the campaign creation response.

2. Cache the catalog response for 15 minutes — reuse within the TTL,
   fetch again after expiry.

3. Add catalog validation before update (PATCH) operations on
   campaigns, ad sets, ads, draft edits, and bulk budget changes.
   Status-only operations (pause/resume/delivery/archive) are excluded.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adds a visible ⛔ CHECKPOINT pattern requiring agents to print a
✅/❌ validation summary for every field checked against the ad
product catalog rules before executing any create or update request.
This makes it impossible for agents to skip validation silently.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When catalog validation finds a failing field, the agent now presents
a recommended fix and asks the user to either accept the recommendation
or provide their own value. No auto-correction — the user always
decides.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adds the response schema for GET /ad_product_catalog to endpoints.md,
showing the create/update/both rule separation per entity type. Updates
all skills to reference the correct rule section based on the operation:
create rules + both for POSTs, update rules + both for PATCHes.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Migrate all ad product catalog fetch and campaign lookup curl commands
to use the api() request wrapper added in #35. This aligns the catalog
validation steps with the rest of the skills.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Codex <noreply@openai.com>
…t-catalog-validation

# Conflicts:
#	tests/test-scenarios.md
Co-Authored-By: Codex <noreply@openai.com>
@amurph491 amurph491 added this to the v1.9.0 milestone Aug 13, 2026
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
@amurph491
amurph491 marked this pull request as ready for review August 14, 2026 13:56
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants