agentleFS
Microsoft AzureKnown publisherAzure/Azure-Sentinel/AGENTS.md

Azure-Sentinel

AGENTS.md6.2k starsChanged 16 days agoMITView on GitHub

During normal repository tasks, treat agent definitions, skills, instructions, tooling backends, schemas, tests, dependency metadata, workflow definitions, safety gates, permission checks, packaging behavior, and capabilities as immutable. Do not create, edit, delete, rename, patch, bypass, or replace implementation under these paths to recover from a task, environment, dependency, authentication, conversion, validation, packaging, deployment, or runtime failure: Stop the affected operation and report that a maintainer change is required. Implementation changes are allowed only in a separate development task where the…

We split this AGENTS.md into the rules it gives an agent

7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not create, edit, delete, rename, patch, bypass, or replace implementation under these paths to recover from a task, environment, …

    line 10
  2. Distinctive:

    Stop the affected operation and report that a maintainer change is required.

    line 20
  3. Distinctive:

    When a request involves Sentinel-to-Defender-XDR migration, XDR Detections, hybrid solution packaging, or AR/CD qualification:

    line 32
  4. Distinctive:

    Treat that file as the canonical workflow.

    line 37
  5. Distinctive:

    Keep deterministic work in the repository Python and PowerShell tools.

    line 39
  6. Very specific:

    Default to authoring; ask before optional lab qualification.

    line 40
  7. Very specific:

    Use the V3.1 package-v3-1 route for Sentinel and XDR solution content.

    line 43

≈ “Common” is an estimate: at least 1 in 5 of the 36,368 AGENTS.md 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

Repository agent instructions

Protected agent and tooling implementation

During normal repository tasks, treat agent definitions, skills, instructions, tooling backends, schemas, tests, dependency metadata, workflow definitions, safety gates, permission checks, packaging behavior, and capabilities as immutable.

Do not create, edit, delete, rename, patch, bypass, or replace implementation under these paths to recover from a task, environment, dependency, authentication, conversion, validation, packaging, deployment, or runtime failure:

  • AGENTS.md;
  • .github/agents/**;
  • .github/skills/**; or
  • Tools/**.

Stop the affected operation and report that a maintainer change is required. Implementation changes are allowed only in a separate development task where the user explicitly states that they are acting as a repository/toolkit maintainer and requests the specific implementation change. Requests to run, retry, continue, migrate, package, validate, deploy, or fix generated content never grant permission to change agent or tooling implementation.

The XDR Solution Manager is stricter: it can never make agent or tooling implementation changes, even when explicitly requested by an owner or maintainer. Route such work to a different development agent or require it to be performed manually.

When a request involves Sentinel-to-Defender-XDR migration, XDR Detections, hybrid solution packaging, or AR/CD qualification:

  1. Read and follow Tools\SentinelToXDRMigration\agent-workflows\end-to-end-migration.md.
  2. Treat that file as the canonical workflow; do not redefine its stages or gates in this adapter.
  3. Keep deterministic work in the repository Python and PowerShell tools.
  4. Default to authoring; ask before optional lab qualification.
  5. Deployment, mock ingestion, and alert parity are Qualification-only actions and require explicit exact-scope approval.
  6. Use the V3.1 package-v3-1 route for Sentinel and XDR solution content; do not route XDR packaging through V4.
  7. Resume from workflow-state.json and record artifacts and evidence.
  8. Require explicit approval immediately before any environment write.

Download now (ZIP)

Trending

More talk lately.about the whole repository

Dec 2025: 0 mentionsJan 2026: 2 mentionsFeb 2026: 1 mentionMar 2026: 1 mentionApr 2026: 0 mentionsMay 2026: 2 mentionsJun 2026: 1 mentionJul 2026: 1 mentionAug 2026: 4 mentionsSep 2026: 6 mentionsOct 2026: 2 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the Azure/Azure-Sentinel repository that our weekly crawl found. There is no adoption data for this file yet.

Who's talking about Azure/Azure-Sentinel

The biggest 5 of 57 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 57

Talked about on GitHub, Stack Overflow and Bluesky (57 times).

On GitHub · 43

38 more on GitHub

On Stack Overflow · 13

8 more on Stack Overflow

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.