Record progress on a task. Provide task_id explicitly. Call this after completing meaningful work, not just at the end — intermediate results are valuable if the session ends or an approach fails. Use result for a human-readable summary of what happened. Use state_patch for structured data (file ...
AI agents use update_task to create or update resources in Conductor — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Conductor environment.
This tool creates or modifies data (task progress, results, notes, goals) in a reversible manner. It does not execute external code, delete data irreversibly, or move financial resources. The ability to rename tasks ('goal to rename the task') and update notes ('can be updated at any point') confirms these are Write operations.
From the tool's definition The tool 'Record progress on a task' with capabilities to update 'result', 'state_patch', 'notes', and 'goal' modifies task state data.
Documented attack patterns abuse exactly the kind of access update_task gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Conductor, and nothing reaches the server without passing your rules. This is the rule we recommend for update_task:
{
"version": "1",
"default": "deny",
"tools": {
"update_task": {
"limits": [
{
"counter": "update_task_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}update_task stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Record progress on a task. Provide task_id explicitly. Call this after completing meaningful work, not just at the end — intermediate results are valuable if the session ends or an approach fails. Use result for a human-readable summary of what happened. Use state_patch for structured data (file paths, counts, flags, API responses) that this task or downstream sibling tasks will need programmatically. Use notes for freeform observations, corrections, or context — unlike result, notes can be updated at any point and cleared by passing null. Use goal to rename the task (pending tasks only). It is categorised as a Write tool in the Conductor MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Conductor MCP server in PolicyLayer and add a rule for update_task: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Conductor. Nothing to install.
update_task is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the update_task rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for update_task. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
update_task is provided by the Conductor MCP server (shannonbay/conductor). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Conductor, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
16 Conductor tools catalogued and risk-classified — across an index of 43,000+ MCP servers.