-
Notifications
You must be signed in to change notification settings - Fork 214
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Added AI Agent Semantic Convention #1739
Conversation
e814811
to
7b7efce
Compare
Signed-off-by: Guangya Liu <gyliu@ibm.com>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Great start!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@lmolkova thanks for all of your comments, most of them are addressed now, but I still need to check how to define the tool datastore
, let me check more for this.
5628dbe
to
9904399
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you, it looks great, left a small comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Personally I think this concentrates too much on remote agents, which AFAIK are not widely used.
I suggest create_agent
should be removed until there's more evidence of the adoption of the underlying pattern.
Hey all, thanks for the comments, we have several follow ups for this PR, I will create related PRs soon to fix. |
### Motivation and Context <!-- Thank you for your contribution to the semantic-kernel repo! Please help reviewers and future users, providing the following information: 1. Why is this change required? 2. What problem does it solve? 3. What scenario does it contribute to? 4. If it fixes an open issue, please link to the issue here. --> Resolves: #10174 This PR instruments Agent invocation operation with traces and adds an example with tracing enabled using Console and Application Insights. Trace operation and attribute names are based on the work in following PR about adding AI Agent Semantic Convention to OpenTelemetry (currently in-progress): open-telemetry/semantic-conventions#1739. Examples from Application Insights: End-to-end transaction:  Traces:  Trace attributes related to Agent:  ### Contribution Checklist <!-- Before submitting this PR, please make sure: --> - [x] The code builds clean without any errors or warnings - [x] The PR follows the [SK Contribution Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md) and the [pre-submission formatting script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts) raises no violations - [x] All unit tests pass, and I have added new tests where possible - [x] I didn't break anyone 😄
### Motivation and Context <!-- Thank you for your contribution to the semantic-kernel repo! Please help reviewers and future users, providing the following information: 1. Why is this change required? 2. What problem does it solve? 3. What scenario does it contribute to? 4. If it fixes an open issue, please link to the issue here. --> Resolves: #10174 This PR instruments Agent invocation operation with traces and adds an example with tracing enabled using Console and Application Insights. Trace operation and attribute names are based on the work in following PR about adding AI Agent Semantic Convention to OpenTelemetry (currently in-progress): open-telemetry/semantic-conventions#1739. Examples from Application Insights: End-to-end transaction:  Traces:  Trace attributes related to Agent:  ### Contribution Checklist <!-- Before submitting this PR, please make sure: --> - [x] The code builds clean without any errors or warnings - [x] The PR follows the [SK Contribution Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md) and the [pre-submission formatting script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts) raises no violations - [x] All unit tests pass, and I have added new tests where possible - [x] I didn't break anyone 😄
### Motivation and Context <!-- Thank you for your contribution to the semantic-kernel repo! Please help reviewers and future users, providing the following information: 1. Why is this change required? 2. What problem does it solve? 3. What scenario does it contribute to? 4. If it fixes an open issue, please link to the issue here. --> Resolves: #10174 This PR instruments Agent invocation operation with traces and adds an example with tracing enabled using Console and Application Insights. Trace operation and attribute names are based on the work in following PR about adding AI Agent Semantic Convention to OpenTelemetry (currently in-progress): open-telemetry/semantic-conventions#1739. Examples from Application Insights: End-to-end transaction:  Traces:  Trace attributes related to Agent:  ### Contribution Checklist <!-- Before submitting this PR, please make sure: --> - [x] The code builds clean without any errors or warnings - [x] The PR follows the [SK Contribution Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md) and the [pre-submission formatting script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts) raises no violations - [x] All unit tests pass, and I have added new tests where possible - [x] I didn't break anyone 😄
### Motivation and Context <!-- Thank you for your contribution to the semantic-kernel repo! Please help reviewers and future users, providing the following information: 1. Why is this change required? 2. What problem does it solve? 3. What scenario does it contribute to? 4. If it fixes an open issue, please link to the issue here. --> Resolves: #10174 This PR instruments Agent invocation operation with traces and adds an example with tracing enabled using Console and Application Insights. Trace operation and attribute names are based on the work in following PR about adding AI Agent Semantic Convention to OpenTelemetry (currently in-progress): open-telemetry/semantic-conventions#1739. Examples from Application Insights: End-to-end transaction:  Traces:  Trace attributes related to Agent:  ### Contribution Checklist <!-- Before submitting this PR, please make sure: --> - [x] The code builds clean without any errors or warnings - [x] The PR follows the [SK Contribution Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md) and the [pre-submission formatting script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts) raises no violations - [x] All unit tests pass, and I have added new tests where possible - [x] I didn't break anyone 😄
### Motivation and Context <!-- Thank you for your contribution to the semantic-kernel repo! Please help reviewers and future users, providing the following information: 1. Why is this change required? 2. What problem does it solve? 3. What scenario does it contribute to? 4. If it fixes an open issue, please link to the issue here. --> Resolves: #10174 This PR instruments Agent invocation operation with traces and adds an example with tracing enabled using Console and Application Insights. Trace operation and attribute names are based on the work in following PR about adding AI Agent Semantic Convention to OpenTelemetry (currently in-progress): open-telemetry/semantic-conventions#1739. Examples from Application Insights: End-to-end transaction:  Traces:  Trace attributes related to Agent:  ### Contribution Checklist <!-- Before submitting this PR, please make sure: --> - [x] The code builds clean without any errors or warnings - [x] The PR follows the [SK Contribution Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md) and the [pre-submission formatting script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts) raises no violations - [x] All unit tests pass, and I have added new tests where possible - [x] I didn't break anyone 😄
### Motivation and Context <!-- Thank you for your contribution to the semantic-kernel repo! Please help reviewers and future users, providing the following information: 1. Why is this change required? 2. What problem does it solve? 3. What scenario does it contribute to? 4. If it fixes an open issue, please link to the issue here. --> Resolves: #10174 This PR instruments Agent invocation operation with traces and adds an example with tracing enabled using Console and Application Insights. Trace operation and attribute names are based on the work in following PR about adding AI Agent Semantic Convention to OpenTelemetry (currently in-progress): open-telemetry/semantic-conventions#1739. Examples from Application Insights: End-to-end transaction:  Traces:  Trace attributes related to Agent:  ### Contribution Checklist <!-- Before submitting this PR, please make sure: --> - [x] The code builds clean without any errors or warnings - [x] The PR follows the [SK Contribution Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md) and the [pre-submission formatting script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts) raises no violations - [x] All unit tests pass, and I have added new tests where possible - [x] I didn't break anyone 😄
### Motivation and Context <!-- Thank you for your contribution to the semantic-kernel repo! Please help reviewers and future users, providing the following information: 1. Why is this change required? 2. What problem does it solve? 3. What scenario does it contribute to? 4. If it fixes an open issue, please link to the issue here. --> Resolves: #10174 This PR instruments Agent invocation operation with traces and adds an example with tracing enabled using Console and Application Insights. Trace operation and attribute names are based on the work in following PR about adding AI Agent Semantic Convention to OpenTelemetry (currently in-progress): open-telemetry/semantic-conventions#1739. Examples from Application Insights: End-to-end transaction:  Traces:  Trace attributes related to Agent:  ### Contribution Checklist <!-- Before submitting this PR, please make sure: --> - [x] The code builds clean without any errors or warnings - [x] The PR follows the [SK Contribution Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md) and the [pre-submission formatting script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts) raises no violations - [x] All unit tests pass, and I have added new tests where possible - [x] I didn't break anyone 😄
- Generate `semconv/v1.31.0` - Stop generating deprecated metric semconv similar to all other generation - Fix acronyms: - `ReplicationController` - `ResourceQuota` ## [`v1.31.0` semantic convention release notes](https://github.com/open-telemetry/semantic-conventions/releases/tag/v1.31.0): <h3>🛑 Breaking changes 🛑</h3> <ul> <li> <p><code>code</code>: <code>code.function.name</code> value should contain the fully qualified function name, <code>code.namespace</code> is now deprecated (<a href="open-telemetry/semantic-conventions#1677" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1677/hovercard">#1677</a>)</p> </li> <li> <p><code>gen-ai</code>: Introduce <code>gen_ai.output.type</code>and deprecate <code>gen_ai.openai.request.response_format</code> (<a href="open-telemetry/semantic-conventions#1757" data-hovercard-type="pull_request" data-hovercard-url="/open-telemetry/semantic-conventions/pull/1757/hovercard">#1757</a>)</p> </li> <li> <p><code>mobile</code>: Rework <code>device.app.lifecycle</code> mobile event. (<a href="open-telemetry/semantic-conventions#1880" data-hovercard-type="pull_request" data-hovercard-url="/open-telemetry/semantic-conventions/pull/1880/hovercard">#1880</a>)<br> The <code>device.app.lifecycle</code> event has been reworked to use attributes instead<br> of event body fields. The <code>ios.app.state</code> and <code>android.app.state</code> attributes<br> have been reintroduced to the attribute registry.</p> </li> <li> <p><code>system</code>: Move CPU-related system.cpu.* metrics to CPU namespace (<a href="open-telemetry/semantic-conventions#1873" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1873/hovercard">#1873</a>)</p> </li> <li> <p><code>k8s</code>: Change k8s.replication_controller metrics to k8s.replicationcontroller (<a href="open-telemetry/semantic-conventions#1848" data-hovercard-type="pull_request" data-hovercard-url="/open-telemetry/semantic-conventions/pull/1848/hovercard">#1848</a>)</p> </li> <li> <p><code>db</code>: Rename <code>db.system</code> to <code>db.system.name</code> in database metrics, and update the values to be consistent with database spans. (<a href="open-telemetry/semantic-conventions#1581" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1581/hovercard">#1581</a>)</p> </li> <li> <p><code>session</code>: Move <code>session.id</code> and <code>session.previous_id</code> from body fields to event attributes, and yamlize <code>session.start</code> and <code>session.end</code> events. (<a href="open-telemetry/semantic-conventions#1845" data-hovercard-type="pull_request" data-hovercard-url="/open-telemetry/semantic-conventions/pull/1845/hovercard">#1845</a>)<br> As part of the ongoing migration of event fields from LogRecord body to extended/complex attributes, the <code>session.start</code> and <code>session.end</code> events have been redefined.</p> </li> </ul> <h3>💡 Enhancements 💡</h3> <ul> <li> <p><code>code</code>: Mark <code>code.*</code> semantic conventions as release candidate (<a href="open-telemetry/semantic-conventions#1377" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1377/hovercard">#1377</a>)</p> </li> <li> <p><code>gen-ai</code>: Added AI Agent Semantic Convention (<a href="open-telemetry/semantic-conventions#1732" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1732/hovercard">#1732</a>, <a href="open-telemetry/semantic-conventions#1739" data-hovercard-type="pull_request" data-hovercard-url="/open-telemetry/semantic-conventions/pull/1739/hovercard">#1739</a>)</p> </li> <li> <p><code>db</code>: Add database-specific notes on db.operation.name and db.collection.name for Cassandra, Cosmos DB, HBase, MongoDB, and Redis, covering their batch/bulk terms and lack of cross-table queries. (<a href="open-telemetry/semantic-conventions#1863" data-hovercard-type="pull_request" data-hovercard-url="/open-telemetry/semantic-conventions/pull/1863/hovercard">#1863</a>, <a href="open-telemetry/semantic-conventions#1573" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1573/hovercard">#1573</a>)</p> </li> <li> <p><code>gen-ai</code>: Adds <code>gen_ai.request.choice.count</code> span attribute (<a href="open-telemetry/semantic-conventions#1888" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1888/hovercard">#1888</a>)<br> Enables recording target number of completions to generate</p> </li> <li> <p><code>enduser</code>: Undeprecate 'enduser.id' and introduce new attribute <code>enduser.pseudo.id</code> (<a href="open-telemetry/semantic-conventions#1104" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1104/hovercard">#1104</a>)<br> The new attribute <code>enduser.pseudo.id</code> is intended to provide a unique identifier of a pseudonymous enduser.</p> </li> <li> <p><code>k8s</code>: Add <code>k8s.hpa</code>, <code>k8s.resourcequota</code> and <code>k8s.replicationcontroller</code> attributes and resources (<a href="open-telemetry/semantic-conventions#1656" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1656/hovercard">#1656</a>)</p> </li> <li> <p><code>k8s</code>: How to populate resource attributes based on attributes, labels and transformation (<a href="open-telemetry/semantic-conventions#236" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/236/hovercard">#236</a>)</p> </li> <li> <p><code>process</code>: Adjust the semantic expectations for <code>process.executable.name</code> (<a href="open-telemetry/semantic-conventions#1736" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1736/hovercard">#1736</a>)</p> </li> <li> <p><code>otel</code>: Adds SDK self-monitoring metrics for span processing (<a href="open-telemetry/semantic-conventions#1631" data-hovercard-type="pull_request" data-hovercard-url="/open-telemetry/semantic-conventions/pull/1631/hovercard">#1631</a>)</p> </li> <li> <p><code>cicd</code>: Adds a new attribute <code>cicd.pipeline.run.url.full</code> and corrects the attribute description of <code>cicd.pipeline.task.run.url.full</code> (<a href="open-telemetry/semantic-conventions#1796" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1796/hovercard">#1796</a>)</p> </li> <li> <p><code>user-agent</code>: Add <code>user_agent.os.name</code> and <code>user_agent.os.version</code> attributes (<a href="open-telemetry/semantic-conventions#1433" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1433/hovercard">#1433</a>)</p> </li> </ul> <h3>🧰 Bug fixes 🧰</h3> <ul> <li><code>process</code>: Fix units of process.open_file_descriptor.count and process.context_switches (<a href="open-telemetry/semantic-conventions#1662" data-hovercard-type="issue" data-hovercard-url="/open-telemetry/semantic-conventions/issues/1662/hovercard">#1662</a>)</li> </ul> --------- Co-authored-by: Robert Pająk <pellared@hotmail.com>
Fixes #1732
Changes
Please provide a brief description of the changes here.
Note: if the PR is touching an area that is not listed in the existing areas, or the area does not have sufficient domain experts coverage, the PR might be tagged as experts needed and move slowly until experts are identified.
Merge requirement checklist
[chore]