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.
npx agentmods add instructions/gblush/impactdotcom-mcp/claude-mdgit clone --depth 1 https://github.com/gblush/impactdotcom-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/instructions/gblush/impactdotcom-mcp/claude-md)<a href="https://agentmods.dev/instructions/gblush/impactdotcom-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/gblush/impactdotcom-mcp/claude-md.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 | $0.01192 | $0.01192 |
| Opus 5 | $0.00596 | $0.00596 |
| Sonnet 5 | $0.00238 | $0.00238 |
| Haiku 4.5 | $0.00119 | $0.00119 |
Grade B, and why
impactdotcom-mcp CLAUDE.md scanned grade B with 1 finding 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 4d 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.
Reads MCP configurationmediumAgent snooping
mcp.json carries server URLs and auth tokens; reading it lets a mod discover and abuse other integrations.
This repo is public. Real `IMPACT_ACCOUNT_SID` / `IMPACT_AUTH_TOKEN` live **only** in the gitignored `.env`. Never put real credentials or PII in any tracked file (`.env.example`, `.mcp.json`, docs use placeholders). Run How it starts
The opening of the file, as written. The whole thing — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
What this is
A local stdio MCP server wrapping the Impact.com partner REST API. Built because Impact.com's hosted MCP can't be used from Claude Code (its OAuth server has no Dynamic Client Registration). Primary purpose: diagnose publisher tracking problems — specifically "clicks but no revenue" — by exposing the click → action → revenue funnel.
Commands
npm run build— compile TypeScript todist/.npm run dev— watch-mode run viatsx.npm run inspect— build + open the MCP Inspector against the server (primary manual test loop).npm run typecheck/npm run lint— must pass; lint enforces no strayconsole.log.npm run check:secrets— run before every commit. Fails if any tracked file contains a real Account SID/Auth Token.
Architecture
src/index.ts— entry; loads.env, validates config, registers tools, connects stdio.src/config.ts— env validation; computesbasePathand Basic-auth header.src/client.ts— the only module that talks HTTP (impactGet,impactGetAll,impactWrite,runAsyncJob). Add new endpoints here, not ad-hocfetchin tools.src/diagnostics.ts— encodes Impact's attribution rules + reversal codes; powersdiagnose_tracking.src/tools/*— one file per domain; each exports aregister*function.tools/index.tswires them up and gates writes.
Non-obvious rules (get these wrong and it breaks)
- stdout is reserved for JSON-RPC. Never
console.log. Log viasrc/logger.ts(log.info/warn/error→ stderr, with secret redaction). - Persona prefix: all paths are under
/{IMPACT_ACCOUNT_TYPE}/{AccountSID}(defaultMediapartners). If every call 403s/404s, the account type is probably wrong (Mediapartners vs Advertisers). - JSON, not XML: the API defaults to XML; the client always sends
Accept: application/json. Don't remove that header. - Tool inputs are
YYYY-MM-DD, but date encoding differs by endpoint (verified live): the/Actionsendpoint rejects bare dates — it needs full ISO‑8601 datetimes (2026-05-27T00:00:00Z), so the Actions tools convert viatoStartOfDayIso/toEndOfDayIso. Reports take bareYYYY-MM-DD. The/Actionsendpoint also has max 45-day range, ~3-year-back limit, default last 7 days, noPage/PageSize(paginate via@nextpageuri), and no OrderId filter — trace an order viafind_action_by_order. - Report params differ per report (verified live): Performance by Brand =
partner_performance_by_program(filtersSTART_DATE/END_DATE); Advanced Action Listing =mp_action_listing_sku(filtersStart Date/End Date, with spaces). Don't assume a uniform param name — callget_report_metadatafirst. get_accounthits the persona root (/Mediapartners/{SID}), not/Accounts— the latter is 403 for scoped Mediapartner tokens. Similarly the single/Clicks/{id}resource may be 403 for a scoped token even whenClickExportworks.- Writes are gated:
create_tracking_linkandimpact_api_requestonly register whenIMPACT_ENABLE_WRITES=true.create_tracking_link/TrackingLinkspath casing is the one write not yet exercised live. - Rate limits are real: the API has an hourly quota and exports (
ClickExport,ReportExport) are expensive — bursting parallel exports exhausts it (retry-aftercan be hours). The client caps concurrency (IMPACT_MAX_CONCURRENCY, default 4) and fails fast on a 429 whose reset is >30s out (reset time included in the error) instead of hanging. Don't loop exports across many days without spacing.
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.
- 4d ago First seen · 48 lines · 1,192 tokens per session scan B 88b7c43322dc
impactdotcom-mcp CLAUDE.md is an instructions file published in the GitHub repository gblush/impactdotcom-mcp (0 stars, last pushed 2mo ago), licensed MIT. It adds 1,192 tokens to every session, about $0.0060 per session on Opus 5. A static security scan graded it B with 1 finding (reads mcp configuration). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.