[atrearyan]

Contribution Type

Please select the track that matches this pull request:

• Bug Fix / Minor Patch (Targeted fix, non-test diff <= 100 lines and < 3 core files in skills/ or agents/; no design doc required)
• Feature / Architectural Change -- Stage 1: Design Doc PR (Refs #___ with design-ready | Adds docs/designs/<name>.md and updates docs/designs/README.md; use Refs instead of Closes so the issue stays open for Stage 2)
• Feature / Architectural Change -- Stage 2: Implementation PR (Closes #___ with pr-ready | Implements merged docs/designs/<name>.md)

Summary of Key Changes

Closes #220.

• README.md line 52: keeps the /egm: prefix instruction for Antigravity (agy CLI) and Claude Code, and adds that agy CLI 1.4.10 completes an unprefixed name to the prefixed command, linking Verify the /egm: prefix in the interactive Antigravity CLI #220. The prefix stays the instruction because README: Antigravity (agy CLI) also requires the /egm: prefix for skills #196 saw unprefixed names fail on 1.4.8.
• distributions/claude/README.md: regenerated with python3 scripts/pack_plugin.py --target claude --output <dir> and replaced with the generated egm-claude folder. Only README.md changed. This also carries over docs: note Antigravity (agy CLI) also requires /egm: prefix #208's earlier README edit, so test_published_claude_runtime_is_current passes. It touches the same line as fix(packaging): sync the packaged readme #209; if fix(packaging): sync the packaged readme #209 merges first, I'll regenerate again.
• Evidence: the interactive results on agy 1.4.10 are in the Verify the /egm: prefix in the interactive Antigravity CLI #220 comment (/egm: menu, /egm-brainstorm, /egm-review, /egm-draft, /egm-nonexistent).

Not changed here: docs/getting_started.md and the website still show unprefixed commands without prefix guidance, TESTING.md's Antigravity row still records 1.4.5, and the Antigravity IDE was not tested.

Human Accountability Affirmation

• I am a human contributor. I have personally reviewed, understood, and can explain every line of this change in review (using AI tools to write code is welcome, but human understanding of the submitted change is required).

Verification

• For Design Doc PRs (Stage 1): Executed /egm-review docs/designs/<name>.md, resolved findings via /egm-resolve, and verified relative links (pytest tests/unit/test_link_integrity.py).
• For Bug Fix / Implementation PRs (Stage 2): Unit tests or test_scenarios.yaml added/updated and pytest tests/unit executed.
• For Harness Support PRs: Verified plugin installation and end-to-end execution of /egm-brainstorm, /egm-draft, and /egm-review on the target harness; attached execution summary or transcript.

Docs-only change; no test change needed. pytest tests/unit: 138 passed.

[n3wth]

Reviewed 1585b99. Both files end up with the same line 52 (the diff shows the same blob, 2d3bae8, for each), which is what test_published_claude_runtime_is_current compares. The packaged copy was also stale from before #208, so regenerating it is the right fix. The #220 comment supports the 1.4.10 behavior, and #196 shows unprefixed names failing on 1.4.8.

One non-blocking wording suggestion inline.

#209 edits the same line in distributions/claude/README.md, so it will conflict once this merges, as the description already notes.

I did not run the unit tests; the 138 passed in the description is the author's run. cla/google and the scope/template check pass; the Tests workflow has not run on this head.

Nothing blocking.

[n3wth]

Re-reviewed at c210baf. The only change since 1585b99 is the wording suggestion from my last round, applied to both files. README.md and distributions/claude/README.md now resolve to the same blob (13e4779), which is what test_published_claude_runtime_is_current compares. The text keeps /egm: as the rule, says 1.4.10 completes an unprefixed name and 1.4.8 does not, and links #220 and #196.

I ran tests/unit locally at this head: 138 passed. The Tests workflow has not run on this head; cla/google and the scope/template check pass.

#209 edits the same line in distributions/claude/README.md, so whichever lands second needs a regenerate, as the description already notes.

No remaining finding.

[atrearyan]

Hi @n3wth , Thanks for the re-review! A few related gaps I noticed while testing, outside this PR's scope:

1. docs/getting_started.md still shows /egm-brainstorm, /egm-draft, and /egm-review with no prefix note. I can send a small follow-up PR that points to the README guidance.
2. TESTING.md's Antigravity row still records 1.4.5. My Verify the /egm: prefix in the interactive Antigravity CLI #220 run only checked invocation, not the full brainstorm/draft/review workflow, so I didn't add a row. I can run the full fixture test on 1.4.10 if you want the record updated.
3. The Antigravity IDE (/plugin install) wasn't tested.

This PR also includes #209's packaged README sync, so #209 shouldn't be needed once this merges.

[n3wth]

Please send the small follow-up PR for docs/getting_started.md. Link to the README guidance for command prefixes. Update the packaged documentation if the source change requires it.

Keep the 1.4.10 invocation result separate from the full workflow results. Do not replace the 1.4.5 workflow record with an invocation-only result.

A full fixture run on 1.4.10 would help. If you run it, record the EGM commit, host version, commands, and results. Keep the Antigravity IDE result unverified until a separate test covers it.