[tjdonley]

Summary

• Add a one-command agentprof demo workflow with bundled synthetic Langfuse traces that produce retry-loop, spec-violation, and multi-agent waste findings.
• Rework the README around the demo, concrete waste findings, local-first positioning, and checked-in example reports.
• Add release/community polish: MIT license, v0.1.0 metadata, changelog, contributing/security docs, issue/PR templates, CI matrix, package build, and wheel smoke checks.
• Add docs/project-pop-spec.md as the execution spec for making AgentProf more visible and compelling.

Validation

• /opt/homebrew/bin/python3.11 -m pytest -> 137 passed.
• /opt/homebrew/bin/python3.11 -m uv sync --frozen --dev.
• /opt/homebrew/bin/python3.11 -m uv run pytest -> 137 passed.
• /opt/homebrew/bin/python3.11 -m uv run agentprof demo --dir /tmp/agentprof-uv-demo.
• /opt/homebrew/bin/python3.11 -m uv build.
• Fresh wheel install smoke: AGENTPROF_HASH_SALT=short /tmp/agentprof-wheel-smoke/bin/agentprof demo --dir /tmp/agentprof-wheel-demo.
• git diff --check.

Notes

• Opened as draft so you can review the public positioning and release/version bump before publishing.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f4018c7cd5

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".