MCP (Model Context Protocol) is an open standard that lets AI agents securely connect to external data sources. Birch MCP is Birch's implementation of that standard - it gives any MCP-compatible agent live access to your automation rules across all platforms and Stage data, so it can both analyze what's running and manage it directly.

With Birch MCP, your AI agent can read your rule logs directly - every check and trigger - and use them to tell you what's actually happening: which rules are firing most, what's conflicting, what's worth refining. It can also report on how much volume is moving through Stage. Instead of digging through logs by hand, you can just ask. Your agent can also act on your rules: it can edit their conditions and tasks, and activate, pause, rename, duplicate, move, and reassign them, across all ad platforms.

Who it's for: Birch MCP is for teams with an established automation setup who want more precision on what's already running - and a faster way to manage it - not a starting point for building rules from scratch.

What your AI agent can do with Birch MCP

Birch MCP covers two kinds of use cases: analyzing your rules and Stage activity, and managing your rules directly.

Analyzing your Rules & Stage activity

Here's what each analysis use case actually gets you:

Use case	What it does	Example prompt
Rule activity & triggers	Shows which rules fire most or which barely act.	“Which rule triggered the most in the last 7 days, and are any of my rules overlapping?” “Which of my rules never fire, and which act on almost every entity each run?”
Rule conditions & gaps	Reviews rule logic to spot missing conditions, refine thresholds, identify overlaps or contradictory actions, and find opportunities for new rules.	“Review all rules on my Meta account. Flag overlaps, contradictory actions, thresholds with no volume guard.” “What conditions is [rule name] checking, and where are the gaps in my setup?”
See what automation actually did	Reviews rule activity to show what fired or what errored.	“What did my rules do last night, what fired and what errored?”
Debugging a decision	Pulls the exact condition values a rule evaluated, so you can see why it paused, activated, or changed a budget.	“Why did rule X pause this ad set? Show the values it checked.”
Find the right metric for a rule	Helps you find metrics available for rule logic.	“What metrics can I use for purchase cost on TikTok?”
See which filters a rule can use	Shows which entity filters are available for a specific platform and entity level.	“What entity filters are available for a Meta ad set rule?”
Peak activity windows	Identifies the times of day or days of the week when most rule-driven changes happen, so you know when to expect activity.	“At what hour of the day do most automated rule changes happen?”
Monitor Stage	Reports how many creatives have been uploaded, including errors, and how many ads have been created through Stage.	“How many creatives were uploaded through Stage in the last 14 days?” “Which creative uploads failed on this account, and which ads got created?”
Answer Birch how-to questions	Searches the Birch Help Center content to answer questions.	“How does the ranking condition work in Birch?”

More things you can ask:

“What's the largest budget change an automated rule made this month?”
“How many ads were paused by automated rules last week?”
“Show me all rules in the [Client A] folder and check for conflicts.”

👉 For prompt examples, see our Birch AI Prompting Guide.

Outside of the supported rule-management actions below, Birch MCP remains read-only. It cannot create, edit, pause, or delete campaigns, ad sets, or ads. For that, see Meta MCP Tools.

Managing your rules

Beyond reading rule activity, your AI agent can manage your rules directly, across all ad platforms and for rules in any state (drafts, paused, and active), except Run a rule now, which works for active rules only. This includes editing rule logic: anything you can set up in the rule editor, your agent can now change for you.

Use case	What it does	Example prompts
Edit rule conditions	Changes the conditions a rule checks: metrics, thresholds, timeframes, and condition groups. Supports all metrics, custom metrics, custom events, and custom timeframes.	“In rule X, raise the CPA threshold from $50 to $60 and only act when there are more than 3 purchases today.” “Switch the ROAS condition in my scaling rule to the last 3 days, excluding today.”
Edit rule tasks	Changes what a rule does when its conditions are met, such as the action type, budget change amount, and limits.	“Change the budget increase in rule X from 20% to 30% and set a max daily budget of $500.” “Make my burnout rule decrease the budget by 25% instead of pausing the ad set.”
Run a rule now	Triggers an immediate check of an active rule, so you don’t have to wait for its next scheduled run. Works on all 4 supported ad platforms, once every 10 minutes per rule. Only works for active rules: drafts and paused rules can’t be run.	“I just updated my ‘Pause high CPA’ rule. Run it now.” “Run my Meta scaling rule now and show me what it changed.”
Duplicate rule	Creates a copy of an existing rule.	“Copy my TikTok budget rule into the [Client B] workspace.” "Duplicate rule X."
Rename rule	Renames a rule.	"Rename rule X to Y."/ “Rename all rules in this folder to the [Client] [Action] [Metric] format.”
Move rules	Moves a rule between folders or subfolders.	"Move rule X to the [Client] folder." “Move all scaling rules into the [Scaling] folder and take the test rules back to the top level."
Turn rules on/off	Enables or disables an existing rule.	“Disable every rule in the [Client A] folder on Meta while we pause the account.”
Set rule ad accounts	Changes which ad accounts a rule runs on.	"Change the ad accounts in scope of rule X to 1, 2, 3."
Change rule filters	Replaces the entity filters that determine which campaigns, ad sets, or ads a rule looks at.	“Make my Meta burnout rule only look at ad sets with ‘Prospecting’ in the name and over 50 impressions in the last 7 days.”
Audit and fix in one go	Combines rule analysis with a management action after you confirm what should be changed.	“Find rules that have not fired in 30 days, list them, then disable the ones I confirm.” “Check my Meta scaling rules for conditions that mix different timeframes, and fix the ones I approve.”

❗To get the newest tools, click Refresh tools in your MCP client. If the new tools still don’t appear, disconnect and reconnect Birch MCP. In Claude, go to Connectors → Birch MCP → Disconnect, then Connect again, and grant write access when the consent screen asks for it.

Getting connected

Setting it up takes under a minute - see Setting up Birch MCP for step-by-step instructions for Claude.ai, ChatGPT, Claude Code, and Codex.

Please feel free to contact our team in the in-app chat bubble or at support@bir.ch in case of any questions!

Birch MCP: Overview & Use Cases

What your AI agent can do with Birch MCP

Analyzing your Rules & Stage activity

Managing your rules

Getting connected

Related Articles