Skip to content

Shorten the ChatGPT server instructions to cross-tool rules - #1040

Merged
helpcodeai merged 2 commits into
mainfrom
helpcodeai/openai-findings-fix
Oct 12, 2026
Merged

helpcodeai merged 2 commits into
mainfrom
helpcodeai/openai-findings-fix

Conversation

@helpcodeai

Copy link
Copy Markdown
Contributor

Summary

OpenAI's plugin scan of our ChatGPT submission flagged the server instructions of the ChatGPT profile on the shared /mcp ("These server instructions appear to execute operations that are not individually exposed"). Those instructions walked through every tool one by one. OpenAI's guidance for server instructions is to carry only rules that span tools, keep the important part in the first 512 characters and not repeat the tool descriptions. This PR does that for the chatgpt profile only.

Cloud and self-hosted: the change only affects connections whose OAuth client registered a chatgpt.com / openai.com redirect URI and only where /mcp serves the fixed tool set (Cloud default, opt-in on self-hosted via MCP_SHARED_ENDPOINT_TOOLS=fixed). The default set served to Claude and every other client, including its instructions, is unchanged.

Changes

  • CHATGPT_INSTRUCTIONS now has three rules: read the workspace guide before first using a connector, confirm each anythingmcp_run_write_tool / anythingmcp_add_connector call, never ask for secrets in the chat (under 512 characters)
  • Tool list, descriptions, schemas and annotations unchanged
  • New spec: ChatGPT instructions stay within 512 characters and do not list the tools; the controller spec compares against the profile's instructions instead of a tool name

Type

  • Bug fix
  • New feature
  • Documentation
  • Refactoring
  • Performance improvement
  • Other (describe below)

Testing

  • Added/updated unit tests
  • Tested manually (describe steps below)
  • No tests needed (explain why)

Run in packages/backend:

  • npx jest src/mcp-server: 22 suites, 336 tests passed
  • npx tsc --noEmit -p tsconfig.json: clean
  • npm run lint: its --fix rewrote two unrelated files, reverted, not part of this PR

The existing spec "leaves the default set exactly as the Claude directory reviewed it" still passes unchanged.

Checklist

  • My code follows the existing code style
  • I have tested my changes locally
  • I have updated documentation (if applicable)
  • All existing tests pass (npm test)
  • This PR has a descriptive title

Related Issues

None.

helpcodeai and others added 2 commits October 12, 2026 09:38
OpenAI's plugin scan flagged the ChatGPT profile's server instructions, which walked through every tool. Following OpenAI's guidance, they now only carry the rules that span tools (read the workspace guide first, confirm every write or connector add, never ask for secrets) in under 512 characters. The tool list, descriptions and annotations are unchanged, and the default set served to Claude and every other client is untouched.
@helpcodeai
helpcodeai merged commit 4306cbf into main Oct 12, 2026
13 checks passed
@helpcodeai
helpcodeai deleted the helpcodeai/openai-findings-fix branch October 12, 2026 07:47
@github-actions github-actions Bot locked and limited conversation to collaborators Oct 12, 2026
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant