Works with the AI Agents you already use
KQL Investigation Planner
Turn an authorized defensive security question into a bounded investigation plan and read-only KQL drafts.
$19
KQL Investigation Planner
Example session with this skill installed
In Sentinel, investigate analyst@example.invalid between 08:00 and 09:00 UTC on January 15, 2026. Available SigninLogs fields: TimeGenerated, UserPrincipalName, ResultType, IPAddress, AppDisplayName, CorrelationId. Draft a read-only first query.
- Read your context and instructions
- Compiled the kql investigation planner
Scope: authorized Sentinel sign-in triage for one synthetic user; end time exclusive. Query draft, unexecuted:
SigninLogs
| where TimeGenerated >= datetime(2026-01-15T08:00:00Z) and TimeGenerated < datetime(2026-01-15T09:00:00Z)
| where UserPrincipalName =~ "analyst@example.invalid"
| project TimeGenerated, ResultType, IPAddress, AppDisplayName, CorrelationId
| order by TimeGenerated asc
| take 200
Verify retention and the supplied schema. Run a separate count if completeness matters; take is a sample cap, not a scan-cost guarantee. Compare source IPs and app context with expected activity. A successful sign-in is an observation, not proof of compromise.
Review limits: this example uses only the supplied evidence. Proposed queries, checks, and follow-up actions remain unexecuted. Confirm the schema, environment, and decision owner before applying any recommendation. No live access, automatic remediation, or guaranteed improvement is included.
Connects securely to your tools. The creator never sees your data.
About this skill
Turn an authorized defensive security question into a bounded investigation plan and read-only KQL drafts.
What you receive
An investigation scope, schema assumptions, staged query drafts, interpretation guidance, and evidence gaps.
When to use it
Plan an authorized defensive investigation and draft read-only KQL using the supplied schema and time range.
What to provide
Supply the relevant code, schema, design notes, brief, or evidence together with the intended business result and constraints. Sanitized examples are sufficient to begin. The agent identifies missing details that materially affect the recommendation and labels conditional conclusions.
Boundaries
This is an advisory workflow for supplied evidence. It does not connect to a database, tenant, production system, or business application by itself. Runtime execution, deployment, remediation, and external actions are not included. No benchmark, security certification, compliance approval, or business outcome is guaranteed. Proposed validation remains unexecuted unless separately run with authorized tools.
Package
One SKILL.md with the complete workflow. One-time purchase. Your AI assistant and any source-system licenses are separate.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Security.
Requires manual schema verification. No live execution or tenant connection. Output is limited to read-only operators. Does not provide remediation or containment actions.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 21 days ago
Nothing to install. It runs in your AI agent.
Running queries is optional and needs your own runtime, licenses and authorized access.
Permissions
Read only user-authorized input files when provided; pasted input is sufficient. Write review artifacts only when requested. No unrestricted network access, credentials, live system changes, or automatic query execution are required.