Feature request: Engram integration with ChatGPT (web, mobile and voice)
Maintainers usually reply within 1 day
Nobody has claimed this yet.
Assessment
- Difficulty
- 5/5
- Estimated time
- Over a week
- Newbie friendliness
- 15/100
Research direction
The issue names only the MCP server area (tools and transport), with no specific files or tests. Start by reading how the existing MCP tools are exposed and what the remote transport allows, then decide whether automatic capture or retrieval is feasible within ChatGPT's tool-calling model. Done means a scoped design proposal from a maintainer, not a code change.
Written by the indexing model from the issue text.
Description
Existing Issue Search
- I searched existing issues and will add evidence to a canonical tracker when applicable.
🔍 Observed Consequence
ChatGPT users cannot reliably use Engram as persistent memory across conversations in the same way they might with supported coding agents.
In our tests, ChatGPT can retrieve existing Engram memories when it invokes a connected MCP tool. However, important decisions, discoveries and project context are not guaranteed to be saved automatically. Relevant memories are also not guaranteed to be retrieved when a new conversation starts.
This makes continuity dependent on the assistant choosing to invoke the right tool or the user explicitly asking it to search or save. The limitation matters for both development and non-development projects.
💡 Desired Outcome
I would like to use Engram as shared, persistent memory directly from ChatGPT—not just from Codex.
Ideally, ChatGPT would retrieve relevant context and save important decisions and findings across conversations as reliably as the platform allows. The experience should cover ChatGPT on the web and in the official mobile applications, including voice interactions where supported.
It would also be useful to access the same project memories from ChatGPT and other supported agents, with clear project boundaries and control over memory writes.
📦 Affected Area
MCP Server (tools, transport)
Release Context
main
🔄 Workarounds and Constraints
We built a custom authenticated remote MCP bridge hosted on Oracle Cloud to connect ChatGPT with Engram Cloud.
With this setup, we have successfully searched and retrieved real project memories created by other agents, including during ChatGPT voice conversations. We also verified retrieval after restarting and renaming our services.
However, retrieval depends on ChatGPT invoking the tool. We have not established reliable automatic capture or persistence of conversations, including voice conversations, or guaranteed context retrieval at the start of a new chat.
We have not found ChatGPT lifecycle hooks equivalent to those used by some coding agents. Tool availability and behaviour may also vary between web, mobile and voice.
📎 Additional Context
Our current setup is:
ChatGPT → authenticated remote MCP bridge on Oracle Cloud → Engram Cloud
Our main goal is to use Engram with ChatGPT as a platform: desktop web, official mobile apps and voice wherever technically supported.
We would appreciate guidance on what can realistically be automated within ChatGPT's current capabilities, as well as secure remote access, project isolation and controlled memory writes.
I'd be happy to share anonymised architectural details and test findings if useful.
Thanks for considering this request!
- Dominant language
- Go
- Stars
- 7.1k
- Forks
- 729
- Avg merge
- 8h 36m
- Merged PRs (30d)
- 281
Getting set up
Starts the project's dev container in your browser, under your own GitHub account.
- No Dockerfile or Docker Compose file
- Has a pull request template
- Read the contributing guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from Gentleman-Programming/engram
-
status:needs-review
Difficulty 2/5 1-3 hours Newbie friendliness 85/100
Gentleman-Programming/engram#2065 · 1 comment ·
Maintainers usually reply within 1 day
-
status:needs-review
Difficulty 4/5 3-5 days Newbie friendliness 38/100
Gentleman-Programming/engram#2278 · 1 comment ·
Maintainers usually reply within 1 day
-
status:needs-review type:feature
Difficulty 4/5 3-5 days Newbie friendliness 48/100
Gentleman-Programming/engram#2243 ·
Maintainers usually reply within 1 day
-
bug(cli): upgrading the binary leaves a stale server and nothing reports the pending restartPossibly taken @dnlrsls claimed this 1 day ago. Openstatus:needs-review type:bug
Gentleman-Programming/engram#2242 · 2 comments · 1 assignee ·
Maintainers usually reply within 1 day
-
bug(claude-code): session binding rewrites mem_save input, so auto mode denies memory writesPossibly taken @dnlrsls claimed this 1 day ago. Openstatus:needs-review type:bug
Gentleman-Programming/engram#2231 · 4 comments · 1 assignee ·
Maintainers usually reply within 1 day
All issues in Gentleman-Programming/engram
Similar issues
-
area/docs framework/hugo help wanted language/markdown
Difficulty 1/5 Under an hour Newbie friendliness 88/100
Maintainers usually reply within 1 day
-
clawsweeper:fix-shape-clear clawsweeper:queueable-fix clawsweeper:source-repro impact:other issue-rating: 🦞 diamond lobster no-stale P2
Difficulty 2/5 1-3 hours Newbie friendliness 75/100
openclaw/crabbox#3060 · 1 comment · 1 reaction ·
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 76/100
kubeflow/semantic-operator#52 ·
Maintainers usually reply within 10 days
-
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
Maintainers usually reply within 1 day
-
enhancement help wanted
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
PramodBasavarajMenasi/siftwise#52 ·
Maintainers usually reply within 1 day