Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/logly/mureonpx agentmods add skills/logly/mureo/_mureo-amazon-adsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/logly/mureo/_mureo-amazon-ads)<a href="https://agentmods.dev/skills/logly/mureo/_mureo-amazon-ads"><img src="https://agentmods.dev/badge/skills/logly/mureo/_mureo-amazon-ads.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00045 | $0.05285 |
| Opus 5 | $0.00023 | $0.02642 |
| Sonnet 5 | $0.00009 | $0.01057 |
| Haiku 4.5 | $0.00005 | $0.00528 |
Grade A, and why
_mureo-amazon-ads scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 347 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Amazon Ads (official MCP, bridged by mureo)
PREREQUISITE: Read
../_mureo-shared/SKILL.mdfor auth, security rules, and output format — and specifically its Plugin platforms (third-party providers) section, which is the contract Amazon rides: honest scope, declaration-first guardrails,action_logpromotion, structural strategy parity for mutations. This file does not repeat any of it.
Canonical platform key:
plugin:mureo-amazon-ads-bridge:amazon_ads. Use it verbatim wherever a mureo surface names a platform — STATE.jsonplatforms,mureo_state_upsert_campaign(platform="plugin:mureo-amazon-ads-bridge:amazon_ads"),mureo_state_platform_metrics_set,action_log,mureo_analytics_*. Amazon's own tools carry no mureo platform key; joining on anything else (amazon,amazon_ads, a tool namespace) silently fails to join.The older
plugin:mureo-amazon-ads-bridgeform (before the key carried the provider, logly/mureo#537) stays valid on read — it resolves to the same platform, and the dashboard still labels it Amazon Ads. If STATE.json already holds an entry under it, keep writing that key: a second entry for one ad account is what double-counts it on the Reports card, and mureo never merges or rewrites either entry.
There is no
mureo amazon-ads …CLI command. The only Amazon CLI ismureo amazon refresh-manifest(rebuilds the tool list). Every operation below is an MCP tool — call it directly.
What this surface is — and is not
- Amazon's own tool names, forwarded verbatim. mureo renames nothing (same treatment as the other official MCPs).
- Naming convention:
<namespace>-<verb>_<resource>— the namespace is separated by a HYPHEN, with underscores inside each part (campaign_management-query_ad,account_management-query_advertiser_account). Native mureo tools are all-underscore (google_ads_campaigns_list), so any name matching, prefix filter, or "which platform is this tool" heuristic must account for the hyphen. - The tool list is the operator's own manifest, written by
mureo amazon refresh-manifestnext tocredentials.jsonand read by the mureo MCP server at start. 85 tools on the reference account. The surface can drift: if a tool named here is absent from the session, say so and suggest re-runningmureo amazon refresh-manifestrather than guessing a name. mureo warns once per process when the served manifest is older than its staleness threshold — read that warning as "everything below may be out of date", and re-check before reporting a capability as missing. - No tool declares an
outputSchema(0 of 85). Response shapes are learned by calling. Only two envelopes are confirmed from live responses:campaign_management-query_campaign→{"campaigns": [...]}andcampaign_management-query_ad→{"ads": [...]}. For everything else, read the actual response before you describe it — never state a field name you have not seen come back. - Analytics are advisory. No mureo analytics module ships for Amazon
(issue #120): anomaly detection,
result_indicatorCV-mismatch, RSA-asset audit and rule-based scoring do not exist here. Reportanalytics_not_available_for_plugin:mureo-amazon-ads-bridge:amazon_adsinstead of inventing a heuristic from the tool schemas. The platform-agnosticanalysis_anomalies_checkdoes apply — see../_mureo-shared/SKILL.md→ Generic anomaly check. - Guardrails are EXACT on the known money tools, with the scan as a floor
under everything. A bridged manifest carries no mureo declarations, so
mureo holds them itself: the 13 money-carrying tools below are declared
by exact argument PATH —
budgets[].budgetValue.monetaryBudgetValue.monetaryBudget.value,budgetCaps.countryMonetaryBudgetSettings.<CC>.value,bid.bid,bid.{baseBid,defaultBid,maxAverageBid}and theirmarketplaceSettings[]overrides — andSTRATEGY.md## Guardrailsbudget/bid caps are enforced on those paths, not inferred. The best-effort pattern scan still runs underneath them, and the larger amount wins, so a money field that drifted shape or that Amazon added after the snapshot falls back to that best-effort cover — a declaration can only ever raise what gets checked, never lower it. That cover is a pattern, not a guarantee: it finds the field when the new name still looks like money (budget/spend/bid, or a plainvalue/amountunder one), so a leaf renamed to something outside that vocabulary is missed exactly as it was before the paths were declared. Any OTHER Amazon tool gets that same pattern scan on its own. An oversized or deeper-than-scannable payload fails closed with a message naming the cause. So: a cap on the tools listed below is exact and quotes the declared path when a value is unreadable; on any other Amazon write call it is strong but not guaranteed, say so in the output, and verify the resulting values after the first mutation.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago Changed 857f4522175f
- 7d ago First seen · 347 lines · 45 tokens per session scan A 9352a0a80ba6
_mureo-amazon-ads is a skill published in the GitHub repository logly/mureo (43 stars, last pushed 4d ago), licensed Apache-2.0. It adds 45 tokens to every session and 5,285 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
adspirer-ads-agent
Adspirer — AI-powered advertising and performance marketing agent. Manage Google Ads, Meta Ads (Facebook & Instagram), LinkedIn Ads, and TikTok Ads via natural language. 100+ tools for paid media campaign creation, live performance analysis, PPC keyword research with real CPC data, budget optimization, ad creative…
adcp-advertising
Automate advertising campaigns with AI. Create ads, buy media, manage ad budgets, discover ad inventory, run display ads, video ads, CTV campaigns, and optimize ad performance. Perfect for marketing automation, programmatic advertising, media buying, ad management, campaign optimization, creative management, and…
google-ads-manager
Google Ads campaign management - campaign setup, keyword research, bid optimization, and performance reporting.
ads-copywriter
Multi-platform ad copy generation for Google Ads, Meta/Facebook, TikTok, LinkedIn with A/B testing variants.
adcp-advertising
Automate advertising campaigns with AI. Create ads, buy media, manage ad budgets, discover ad inventory, run display ads, video ads, CTV campaigns, and optimize ad performance. Perfect for marketing automation, programmatic advertising, media buying, ad management, campaign optimization, creative management, and…
adspirer-ads-agent
Adspirer — AI-powered advertising and performance marketing agent. Manage Google Ads, Meta Ads (Facebook & Instagram), LinkedIn Ads, and TikTok Ads via natural language. 400+ tools for paid media campaign creation, live performance analysis, PPC keyword research with real CPC data, budget optimization, ad creative…