Skip to content

Restore official HyperMemory skill and streamline lifecycle - #2

Closed
i-am-dogboy wants to merge 1 commit into
mainfrom
codex/hypermemory-fire-and-forget
Closed

Restore official HyperMemory skill and streamline lifecycle#2
i-am-dogboy wants to merge 1 commit into
mainfrom
codex/hypermemory-fire-and-forget

Conversation

@i-am-dogboy

@i-am-dogboy i-am-dogboy commented Aug 17, 2026

Copy link
Copy Markdown
Member

Summary

  • restore the official HyperMemory MCP skill baseline, including its YAML metadata and full protocol sections
  • classify only narrow standalone greetings and acknowledgements as lightweight so prompts like hey skip recall without suppressing meaningful short requests
  • dispatch one fresh memory-writer with fork_turns="none" and continue immediately without waiting, polling, reading, or messaging it
  • align packaged agent contracts, repository guidance, documentation, and regression tests with the fire-and-forget lifecycle
  • preserve the token-accounting fixes already merged to main

Validation

  • .venv/bin/ruff check plugins packages tests scripts
  • .venv/bin/pytest -q (19 passed)
  • .venv/bin/python scripts/build_plugin_archives.py
  • HyperMemory and HyperColab plugin validators
  • HyperColab generic skill validator
  • python -m py_compile for the HyperMemory hook and token listener

The official HyperMemory skill intentionally retains version, enforcement, and trigger; the generic Codex skill validator rejects those official extended fields, so the plugin validator and repository tests cover it instead.

@i-am-dogboy
i-am-dogboy deleted the codex/hypermemory-fire-and-forget branch August 17, 2026 08:41
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant