fix: restrict bonk commands to repo members, collaborators, and owners

[james-elicx]

Summary

• Adds author_association check to both bonk.yml and bigbonk.yml, matching the existing pattern in deploy-preview-command.yml
• /bonk, @ask-bonk, and /bigbonk can now only be triggered by MEMBER, COLLABORATOR, or OWNER — preventing arbitrary users from invoking the AI agent

[james-elicx]

/bigbonk review this pr

[pkg-pr-new]

Open in StackBlitz

npm i https://pkg.pr.new/vinext@327

commit: 1be38ad

[github-actions]

Example	Preview	Production	Original
app-router-cloudflare	preview	production
pages-router-cloudflare	preview	production
app-router-playground	preview	production	original
realworld-api-rest	preview	production
nextra-docs-template	preview	production
benchmarks	preview	production
hackernews	preview	production	original

[ask-bonk]

This PR is straightforward and correct. Here's my review:

Changes look good. Both bonk.yml and bigbonk.yml now gate on author_association, matching the exact same pattern already established in deploy-preview-command.yml (lines 22-26). This closes a real security gap — without this check, any GitHub user could comment /bonk or /bigbonk on an issue or PR and trigger the AI agent workflow, which has contents: write and pull-requests: write permissions and access to secrets (Cloudflare API tokens).

A few specific observations:

1. Pattern consistency — The author_association values (MEMBER, COLLABORATOR, OWNER) and the multi-line YAML formatting match deploy-preview-command.yml exactly. Good.

2. Existing bot check preserved — The github.event.sender.type != 'Bot' guard remains, preventing recursive triggers if the agent posts comments containing the trigger words.

3. No functional regressions — The trigger conditions (/bonk, @ask-bonk, /bigbonk) are unchanged; only the authorization gate is added.

4. Minor note — contains() is a substring match, so a comment like "don't use /bonk" would still trigger the workflow for authorized users. This is pre-existing behavior and not introduced by this PR, so not a concern here.

No issues found. This is a clean security hardening change.

github run