Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/OrbiAds/Orbiads-GAM-MCPWrote 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/commands/orbiads/orbiads-gam-mcp/adops-deal)<a href="https://agentmods.dev/commands/orbiads/orbiads-gam-mcp/adops-deal"><img src="https://agentmods.dev/badge/commands/orbiads/orbiads-gam-mcp/adops-deal/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/orbiads/orbiads-gam-mcp/adops-deal"><img src="https://agentmods.dev/badge/commands/orbiads/orbiads-gam-mcp/adops-deal.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00028 | $0.01333 |
| Opus 5 | $0.00014 | $0.00666 |
| Sonnet 5 | $0.00006 | $0.00267 |
| Haiku 4.5 | $0.00003 | $0.00133 |
Grade A, and why
adops-deal 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 11d 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 — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GAM Programmatic Deals
Always confirm the tenant first: get_my_tenant_id. Write operations require a confirmation_token from a prior preview call.
Naming conventions (always fetch before creating)
Before naming any proposal or deal, call settings(action="get_naming_conventions") and apply the returned pattern. Convention violations cause ambiguity in billing reports. If no pattern is configured, ask the user to confirm the name explicitly.
Estimating cost before writes
For any proposal, run deals(action="estimate_deal_cost", params={deal_type, targeting, impressions_goal, cpm_floor}) first — it returns a credit breakdown at zero cost. Surface the estimate to the user before proceeding.
list
Calls deals(action="list_deals") and deals(action="list_auctions") in parallel. Render a table: type | name | buyer | status | CPM floor.
get <deal-id>
Call deals(action="get_deal", params={deal_id}) for PMP deals or deals(action="get_proposal", params={proposal_id}) for Marketplace proposals. Show full targeting, floor price, buyers, and associated line items.
For auction details: deals(action="get_auction", params={auction_id}).
create
Ask the user for deal type, then follow the corresponding path:
PMP deal or private auction:
- Fetch available buyers:
deals(action="list_buyers"). - Fetch naming conventions:
settings(action="get_naming_conventions"). - Estimate cost:
deals(action="estimate_deal_cost", params={...}). - Call
deals(action="create_deal", params={name, buyer_id, cpm_floor, targeting, ...})— preview step (0 cr, no token required). - For private auctions:
deals(action="create_auction", params={name, floor_price, targeting, ...}).
Marketplace PG/PD proposal (costs 5 cr, token required):
- Validate feasibility first:
deals(action="adcp_validate", params={targeting, goal, cpm})— free. - Preview the proposal:
deals(action="adcp_preview", params={...})— returns estimated delivery. Free. - Fetch buyers:
deals(action="list_buyers"). - Fetch naming conventions:
settings(action="get_naming_conventions"). - Estimate credit cost:
deals(action="estimate_deal_cost", params={...}). - Preview ADCP:
deals(action="adcp_create", params={..., dry_run: true})— returns a signedExecutionPlanwithdealSpecs[],packageCount,partialFailureRisk,estimatedCost, andconfirmationToken. Show the full plan to the user. IfpartialFailureRiskis present or high, call out that package 0..N-1 may remain in GAM if package N fails. - Execute ADCP:
deals(action="adcp_create", params={..., confirmation_token: "<token>"})— creates throughDealPipeline(PMP REST / PG SOAP / PD SOAP). - Add line items when needed:
deals(action="create_proposal_line_items", params={proposal_id, ...}). Costs 3 cr, requires a freshconfirmation_token.
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.
- 11d ago First seen · 91 lines · 28 tokens per session scan A 0f2f39345094
adops-deal is a command published in the GitHub repository OrbiAds/Orbiads-GAM-MCP (3 stars, last pushed 1mo ago), licensed MIT. It adds 28 tokens to every session and 1,333 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
verify-pr
Verify a PR's frontend changes through browser automation.
explore
Explore a web page using browser automation.
record
Record browser actions into a test definition.
test-init
Initialize qa-use test directory with example test.
test-run
Run E2E tests with qa-use CLI.
context-resume
Call the context MCP tool with action: "resume", project: "$ARGUMENTS" (if no argument given, infer the project name from the current working directory name), and rootPath: " ".