Chronicle: Surface agent_name, fix SQL truncation, and improve query reliability

[vijayupadya]

Surface agent_name, fix SQL truncation, and improve query reliability

[github-actions]

Screenshot Changes

Base: 70d194f9 Current: 794b223e

Changed (3)

chat/aiCustomizations/aiCustomizationManagementEditor/McpBrowseMode/Light

Before	After

agentSessionsViewer/WithBadge/Dark

Before	After

agentSessionsViewer/WithBadge/Light

Before	After

[Copilot]

Pull request overview

This PR updates Copilot Chronicle session storage and querying so sessions can surface an agent_name, SQL query results are less aggressively truncated, and chronicle prompts/queries use more reliable session context.

Changes:

• Add agent_name and agent_description fields to the Chronicle sessions schema and SessionRow typing; populate agent_name from OTel span attributes.
• Increase chronicle query limits (to 100) and switch sessions queries to SELECT * to include the new columns.
• Improve session_store_sql result formatting by using an adaptive per-cell truncation limit instead of a fixed 100 chars.

Show a summary per file

File	Description
extensions/copilot/src/platform/chronicle/node/sessionStore.ts	Bumps schema version and adds DB migration + upsert support for agent_name/agent_description.
extensions/copilot/src/platform/chronicle/common/sessionStore.ts	Extends SessionRow interface with agent_name and agent_description.
extensions/copilot/src/extension/tools/node/sessionStoreSqlTool.ts	Makes SQL output truncation adaptive using a global formatting budget.
extensions/copilot/src/extension/intents/node/chronicleIntent.ts	Updates cloud sessions query/limits and schema description; plumbs agent fields from cloud rows.
extensions/copilot/src/extension/chronicle/vscode-node/sessionStoreTracker.ts	Persists agent_name into the local session store during session initialization.
extensions/copilot/src/extension/chronicle/common/standupPrompt.ts	Includes agent in standup session lines and uses first turn text as a fallback “summary”.

Copilot's findings

• Files reviewed: 6/6 changed files
• Comments generated: 2

[uditkotecha1-hub]

🕵️ CodeSherlock Review:

🔴 Critical Issues:

• The change in the SQL queries from selecting specific columns to using SELECT * may lead to performance degradation and should be avoided without a good reason, especially when working with large datasets. This can result in retrieving unnecessary data and increase latency.

🟡 Improvements:

• The handling of session summaries has been improved by using the first turn's user message as a fallback, which enhances the completeness of the prompt and the overall usability of the data.
• The addition of the agent_name to the session structure increases the context of the sessions for better tracking and understanding.

🟢 Good Practices:

• The implementation of default values (e.g., 'unknown' for branch and repository) when data is missing follows a good defensive programming strategy, improving robustness.
• Clearly documented SQL query rules and guidelines within the comments ensure maintainability and understanding of the codebase.

💡 Suggestions:

• Consider returning only the necessary columns in the SQL queries, as this is a common best practice for readability and performance.
• If the increase in the LIMIT in the queries is essential, ensure that the corresponding logic to handle larger datasets is in place to avoid performance issues.
• Add comments to clarify the importance of using JOIN between sessions and turns, especially in the context of employing session summaries.

Risk Level: MEDIUM