Skip to content

Run persona evaluations through Claude Code sub-agents - #65652

Merged
pelikhan merged 3 commits into
mainfrom
copilot/agent-persona-exploration
Oct 4, 2026
Merged

pelikhan merged 3 commits into
mainfrom
copilot/agent-persona-exploration

Conversation

Copilot AI commented Oct 4, 2026 •

Copy link
Copy Markdown
Contributor

The persona explorer’s CLI/MCP interface could not invoke freeform custom-agent prompts, leaving scenarios unevaluated. Route evaluations through Claude Code and preserve invocation failures separately from quality scores.

  • Invocation: Use an inline Claude Code evaluator following the repository’s agentic-workflows guidance.
  • Baseline: Re-run the same four scenarios (PM-1, DS-1, LC-1, LC-2) once; resume persona rotation after all four invocations succeed.
  • Results: Store structured recommendations, five rubric scores, and invocation status/error in cache memory.
engine: claude

Copilot AI linked an issue Oct 4, 2026 that may be closed by this pull request
Copilot AI and others added 2 commits October 4, 2026 20:45
Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Copilot AI changed the title [WIP] Explore agent persona for program manager, designer, legal compliance Run persona evaluations through Claude Code sub-agents Oct 4, 2026
Copilot AI requested a review from pelikhan October 4, 2026 20:51
@pelikhan
pelikhan marked this pull request as ready for review October 4, 2026 20:58
Copilot AI balanced review requested due to automatic review settings October 4, 2026 20:58

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot review overview

🟡 Changes recommended

The schema change rejects valid expression-based concurrency queue values covered by an existing test.

Review effort: Balanced
Findings: 1 High severity

Open (1)
What changed in this PR

Routes persona evaluations through an inline Claude Code sub-agent and adds structured baseline tracking.

Changes:

  • Switches the workflow from Codex to Claude.
  • Adds four-scenario baseline execution and structured cached results.
  • Regenerates the workflow lock file.
File Description
.github/​workflows/​agent-persona-explorer.md Defines baseline evaluation and inline sub-agent behavior.
.github/​workflows/​agent-persona-explorer.lock.yml Regenerates execution for Claude Code.
pkg/​workflow/​schemas/​github-workflow.json Narrows concurrency queue validation.

💡 Add a code-review agent skill for context-aware, tailored reviews. Learn more in the docs.

Comment on lines +45 to +46
"type": "string",
"enum": ["single", "max"],
@pelikhan
pelikhan merged commit 7dbbced into main Oct 4, 2026
35 checks passed
@pelikhan
pelikhan deleted the copilot/agent-persona-exploration branch October 4, 2026 21:05
@github-actions

github-actions Bot commented Oct 5, 2026

Copy link
Copy Markdown
Contributor

🎉 This pull request is included in a new release.

Release: v0.91.0

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Agent Persona Exploration - 2026-10-04

3 participants