gooseworks-ai/goose-skills / skills/shared/ad-management-intake
ad-management-intake
UnverifiedThe intake stage of a Meta ad harness. Researches the brand first, states what it assumes, and asks only what it cannot infer and what would change money, claims or legal exposure if guessed wrong — at most four questions a turn, in plain business language. Writes the brand's stable facts (ads/brand.md), the campaign brief, and a log of what was asked vs assumed. Use it when someone wants to run Meta ads and the brand's ad facts are not yet recorded.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 17.3 KB |
| agents/openai.yaml | config | 242 B |
| references/business-types.md | markdown | 3.4 KB |
| references/gooseworks-adapter.md | markdown | 7.0 KB |
| references/local-adapter.md | markdown | 2.4 KB |
| skill.meta.json | config | 745 B |
references/local-adapter.md
View on GitHub# Local adapter
## Summary
Use this mode when there is no GooseWorks: a host with a direct Meta connection, or planning only.
Intake never touches Meta, so both work the same way. Research comes from the brand's own pages,
and everything is written to a local `ads/` folder.
## Research (the skill's "research step")
- If the user gave you notes or documents about the brand, read those first. They count as
research; label facts from them "(per <name>, brand notes)".
- Then, if the host has web tools, read the brand's own site: the home page, the pricing page,
the product pages, the about page. Nothing more: no broad market study.
- With no notes and no web tools, there is no research: treat the brand as cold (up to 20
questions in total, still at most four a turn).
- Say which notes or pages you read when you state your assumptions.
## Questions
Use the host's own question control (in Claude Code, AskUserQuestion: 1–4 questions, 2–4
options). Without one, ask in plain numbered text, still at most four a turn.
## The `ads/` docs
- A local `ads/` folder in the working directory the user chose for this brand. Create it on the
first write.
- Write `local` for `brand_id` and `coworker_agent_id`.
- Whole-file writes with the host's file tools. Check the folder with the `meta-ad-manager`
skill's validator, run from the directory that holds `ads/`:
`python3 <meta-ad-manager folder>/contract/validate.py ads`.
## The brief lives in `strategy.md`
There is no campaign record outside GooseWorks, and the contract allows no extra files in `ads/`.
So the brief's facts go into the campaign's `strategy.md` **v1 stub**, each labelled as the
Output section says. This is the one exception to "if `strategy.md` exists, leave it alone":
intake owns these brief lines in local mode.
| Brief fact | `strategy.md` section |
|---|---|
| goal (what the money should produce), offer and price, destination, constraints | `## Goal`, as short labelled lines (`Offer:`, `Destination:`, `Constraints:`) |
| audience | `## Audience` |
| total budget and run time (the brief's budget and date fields) | `## Budget` |
| the success number | `## Success metric` |
Every other section stays "Not decided yet — the strategy stage sets this." A correction to one
of these facts rewrites that line and appends the `Corrected:` entry to `decisions.md`. If the
strategy was already approved, the next strategy run treats the correction as a re-read.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |