MicrosoftKnown publishermicrosoft/copilot-agent-inventory-samples/samples/agent-warden/skills/bulk-action-plan/SKILL.mdbulk-action-plan
Skill29 starsChanged 29 days agoMITView on GitHub
Turn a broad governance intent (e.g. "block every agent shared with the whole org that hasn't been reviewed" or "move all Finance agents to the Finance group") into a reviewable action plan file the admin approves BEFORE anything is executed. Use when the admin wants to act on many agents at once and needs to check the plan first.
In this repository2 other files it gives its agents.
- change-logSkill
- governance-reportSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 33
Only after explicit approval, execute the approved rows one action at a time through the existing functions, following the confirmation and …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
bulk-action-plan
When to use
Trigger when a request implies a change across multiple agents at once — "block all …", "reassign every … to …", "clean up the agents that …". For a single, unambiguous agent, use the existing action functions directly instead.
Inputs
- The
Agentsinventory (knowledge source) to resolve which agents match the intent. - The API plugin to read current state (
warden_GetAgentDetail) and to resolve people/groups (warden_SearchUsers,warden_SearchGroups) so the plan shows names, not just ids.
What it does (runs code)
- Interpret the intent and select the matching agents from the inventory.
Using CodeInterpreter, write
action-plan.csvwith one row per proposed change:Column Meaning AgentIdThe agent's Agent IdNameDisplay name Actionblock/unblock/reassign-availability/reassign-deploymentCurrentStateResolved before-state (blocked flag, or current audience names) ProposedStateResolved after-state ApprovedLeft blank for the admin to mark yes/no- Present the plan and ask the admin to review and edit it (they can change
Approved, remove rows, or re-upload an edited file). - Only after explicit approval, execute the approved rows one action at a time through the existing functions, following the confirmation and write-back rules, and record each result (hand off to the
change-logskill).
Outputs
action-plan.csv(created for review; the admin may edit and return it).
Notes / constraints
- The plan file is a proposal. Generating it performs no writes.
- High-impact actions (block, removing everyone's access) must still be called out explicitly and confirmed.
- Reassignments replace the whole audience collection, so
ProposedStatemust show the full intended set, not just additions.
Status
Draft stub — not yet referenced by the agent.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.