fix: tea agent start

[muratkeremozcan]

What

This restores the TEA conversational agent as a native installable entrypoint and renames it to bmad-tea for consistent end-user invocation across BMAD surfaces. It also updates BMAD’s manifest generation so native SKILL.md agent directories are exported correctly to skill-format IDEs like Codex and Claude Code.

Why

TEA workflows were installing, but the dedicated Test Architect chat/menu entrypoint was being skipped because BMAD only treated native SKILL.md directories as installable when the manifest type was skill, not agent.

Renaming the entrypoint to bmad-tea also makes the conversational agent name consistent and easier for users to discover.

How

• Updated BMAD manifest generation to treat native SKILL.md directories with manifest type agent the same way as native skills for verbatim IDE installation, while preserving agent metadata collection.
• Added regression coverage in BMAD installer tests for native type: agent SKILL.md entrypoints and module scanning.
• Renamed the TEA conversational agent from tea-agent-testarch to bmad-tea in source, tests, and docs, and documented $bmad-tea as the Codex conversational entrypoint.

Testing

Verified the BMAD installer behavior with node test/test-installation-components.js in BMAD-METHOD. Verified the TEA module structure and renamed entrypoint with node test/test-installation-components.js in bmad-method-test-architecture-enterprise.

[augmentcode]

🤖 Augment PR Summary

Summary: Restores the TEA conversational agent as a native installable entrypoint and standardizes end-user invocation as bmad-tea.
Changes:

• Manifest generation treats native SKILL.md directories with type: agent like skills for verbatim IDE export while preserving agent metadata collection.
• Module scanning now recognizes modules that contain only native entrypoints via type: skill/type: agent manifests.
• Installer tests add regression coverage for agent-type SKILL.md entrypoints, module detection, and skill-manifest.csv output.

Technical Notes: Supports skill-format IDEs (e.g., Codex/Claude Code) by exporting native agent entrypoint directories without frontmatter transformation.

_{🤖 Was this summary useful? React with 👍 or 👎}

[augmentcode]

Review completed. 1 suggestion posted.

Comment augment review to trigger a new review at any time.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 9dffa82f-a90a-4db9-8ca7-d69f0d6f953d

📥 Commits

Reviewing files that changed from the base of the PR and between 4cb58ba and e06fa76.

📒 Files selected for processing (3)

• test/test-installation-components.js
• tools/cli/installers/lib/core/manifest-generator.js
• tools/cli/installers/lib/ide/_config-driven.js

---

📝 Walkthrough

Walkthrough

This PR extends manifest generation and skill collection logic to recognize type: agent manifests as native SKILL.md entrypoints alongside type: skill manifests. Tests verify agent entrypoints appear in both skills and agents collections, and documentation is updated to reflect unified native-entrypoint handling.

Changes

Cohort / File(s)	Summary
Test Fixtures & Assertions test/test-installation-components.js	Added type: agent fixture directory with agent manifest (bmad-skill-manifest.yaml and SKILL.md) and extended Suite 29 assertions to verify agent entrypoints are emitted into both generator29.skills[] and generator29.agents[]. Added nested agent-only module coverage and manifest CSV verification.
Manifest Generator Refactoring tools/cli/installers/lib/core/manifest-generator.js	Added isNativeSkillDirType() helper to treat both 'skill' and 'agent' as native SKILL.md entrypoints. Updated collectSkills(), _hasSkillManifestRecursive(), and scanInstalledModules() to use this helper and recognize type: agent manifests. Adjusted warning/debug logging semantics to reference "native SKILL.md entrypoints".
Documentation Update tools/cli/installers/lib/ide/_config-driven.js	Updated installVerbatimSkills docstring to describe installed content as "verbatim native SKILL.md directories" for consistency.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~13 minutes

Possibly related PRs

• BMAD-METHOD#2035 — Flips many manifests from agent to skill type, directly related to the type-handling changes in this PR
• BMAD-METHOD#1857 — Previously added recursive detection of type: skill subdirectories in manifest-generator; this PR extends that to include type: agent
• BMAD-METHOD#1868 — Modified manifest-generator to treat SKILL.md as canonical entrypoint; this PR extends native-entrypoint handling to include type: agent alongside type: skill

Suggested reviewers

• alexeyv
• pbean
• bmadcode

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 inconclusive)

Check name	Status	Explanation	Resolution
Title check	❓ Inconclusive	The title 'fix: tea agent start' is vague and does not clearly convey the main objective of the PR, which involves renaming and enabling the TEA agent as an installable entrypoint and updating manifest generation for agent-type SKILL.md directories.	Consider a more descriptive title such as 'fix: enable native SKILL.md agent directories and rename tea-agent-testarch to bmad-tea' to better reflect the core changes.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description check	✅ Passed	The pull request description comprehensively explains the what, why, and how of the changes, directly relating to the implementation and testing of the agent entrypoint restoration and manifest generation updates.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/tea-agent-start

📝 Coding Plan

• Generate coding plan for human review comments

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

Tip

CodeRabbit can use oxc to improve the quality of JavaScript and TypeScript code reviews.

Add a configuration file to your project to customize how CodeRabbit runs oxc.