test(agent-service): cover prompt composition

[aglinxinyuan]

What changes were proposed in this PR?

Adds isolated unit coverage for agent system-prompt composition from workflow metadata. The tests exercise allowlist filtering, language-specific UDF guidance, the default all-operator path, and the missing-operator fallback.

Any related issues, documentation, discussions?

Closes #7187

How was this PR tested?

• bun test src/agent/prompts.spec.ts — 4 passed.
• bun run typecheck
• bunx prettier --check "src/**/*.{ts,js,json}"

Mutation proof (each production change was reverted):

Production mutation	Focused test result
Inverted non-empty allowlist selection	3 failed
Omitted operator descriptions	1 failed
Changed the no-operators fallback message	1 failed

Was this PR authored or co-authored using generative AI tooling?

Generated-by: Codex (GPT-5)

[github-actions]

Automated Reviewer Suggestions

Based on the git blame history of the changed files, we recommend the following reviewers:

• No candidates found from git blame history.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 80.65%. Comparing base (348fa88) to head (4095152).
⚠️ Report is 6 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff              @@
##               main    #7190      +/-   ##
============================================
+ Coverage     80.58%   80.65%   +0.07%     
  Complexity     3890     3890              
============================================
  Files          1162     1162              
  Lines         46264    46259       -5     
  Branches       5153     5153              
============================================
+ Hits          37280    37309      +29     
+ Misses         7319     7285      -34     
  Partials       1665     1665              

Flag	Coverage Δ		*Carryforward flag
access-control-service	70.00% <ø> (ø)		Carriedforward from 66afe8d
agent-service	78.49% <ø> (+1.07%)	⬆️
amber	75.59% <ø> (ø)		Carriedforward from 66afe8d
computing-unit-managing-service	20.49% <ø> (ø)		Carriedforward from 66afe8d
config-service	65.97% <ø> (ø)		Carriedforward from 66afe8d
file-service	66.80% <ø> (ø)		Carriedforward from 66afe8d
frontend	83.22% <ø> (ø)		Carriedforward from 66afe8d
notebook-migration-service	78.89% <ø> (ø)		Carriedforward from 66afe8d
pyamber	97.36% <ø> (ø)		Carriedforward from 66afe8d
workflow-compiling-service	26.31% <ø> (ø)		Carriedforward from 66afe8d

*This pull request uses carry forward flags. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[mengw15]

LGTM

[Copilot]

Pull request overview

Adds Bun unit tests in the agent-service to validate how the system prompt is composed from workflow operator metadata, including allowlist filtering, compact schema rendering, UDF guidance sections, and the “no operators available” fallback. This fits into the agent-service prompt layer by locking in expected prompt composition behavior independent of backend availability.

Changes:

• Add isolated buildSystemPrompt coverage for allowlist vs. default-all operator selection.
• Verify language-specific UDF guidance inclusion/exclusion behavior.
• Cover the missing-operator allowlist fallback (“No operators available.”).

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.