Skip to content

[Resource]: #768

@ErebusEnigma

Description

@ErebusEnigma

Display Name

context-memory

Category

Agent Skills

Sub-Category

General

Primary Link

https://github.com/ErebusEnigma/context-memory

Author Name

ErebusEnigma

Author Link

https://github.com/ErebusEnigma

License

MIT

Other License

No response

Description

Persistent, searchable context storage across Claude Code sessions using SQLite + FTS5. Saves sessions with AI-generated structured summaries, topic extraction, and code snippet preservation. Two-tier full-text search returns results in under 50ms. Includes auto-save hooks, pre-compaction checkpoints, a web dashboard with analytics, and an MCP server for programmatic access. 364 tests across 12 modules, zero core dependencies, cross-platform.

Validate Claims

364 tests passing across 12 modules on Python 3.8, 3.11, and 3.12 via GitHub Actions CI. Install with git clone https://github.com/ErebusEnigma/context-memory && cd context-memory && python install.py. Test with /remember "test session" then /recall test to verify round-trip.

Specific Task(s)

Save a Claude Code session to persistent memory and search for it in a new session. Thereby remembering past context/important information.

Specific Prompt(s)

  1. /remember "Implemented user authentication with JWT" 2. In new session: /recall authentication

Additional Comments

Everyone reaches for fancier retrieval (vectors, embeddings, re-rankers) when the real issue is that the data wasn't stored well in the first place. If your summaries are rich and descriptive, even basic keyword search finds what you need. If your summaries are thin, no amount of vector math will recover what was never captured.

Recommendation Checklist

  • I have checked that this resource hasn't already been submitted
  • My resource provides genuine value to Claude Code users, and any risks are clearly stated
  • All provided links are working and publicly accessible
  • I am submitting only ONE resource in this issue
  • I understand that low-quality or duplicate submissions may be rejected

Metadata

Metadata

Assignees

No one assigned

    Labels

    resource-submissionThis Issue submits a new resource to the listvalidation-passedResource has passed initial validation

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions