Skip to content
Slop Bucket

Updated daily · last run Nov 21, 07:02 UTC

Trending in the bucket

Every AI dev tool the internet posted about, ranked by who's talking about it. Sorted. Mostly.

In the bucket
3.1K
Posts read
43K
Tossed out
33.7K

119 tools, ranked by sources, chatter and star growth. How ranking works

  1. #1

    deadweight

    Memory & context

    Rust CLI that measures token overhead in a Claude Code setup from plugins, skills, MCP servers and hooks, and disables unused ones.

    GHHN#claude-code#cli#macos2 posts · 2 days ago★ 15

  2. #2

    Headroom

    Memory & context

    Local context compression layer that shrinks AI agent prompts (tool outputs, logs, files) to cut token usage, with proxy, CLI, and MCP serve

    GHHN#claude-code#codex#cursor7 posts · 11 days ago★ 80.2K +611/wk

  3. #3

    GBrain

    Memory & context

    MIT-licensed memory brain for AI agents: stores facts with sources, retrievable over MCP by Claude Code, Codex, Cursor and others; self-host

    GH#claude-code#codex#mcp-server1 post · 11 days ago★ 33.4K +472/wk

  4. #4

    rtk

    Memory & context

    Rust CLI that filters and compresses command output before it reaches a coding agent's context, reducing token consumption.

    HN#claude-code#codex#cursor1 post · 15 days ago★ 87.3K +801/wk

  5. #5

    JetBrains Context

    Memory & context

    A repository intelligence layer that indexes codebases and provides semantic retrieval for coding agents like Claude Code and Codex.

    HN#claude-code#codex#vscode2 posts · 10 days ago

  6. #6

    memsaga

    Memory & context

    Git history and two-way sync for Claude Code memory, CLAUDE.md, skills and config files, with commits named after each session.

    HN#claude-code#macos#linux2 posts · 11 days ago★ 23 +2/wk

  7. #7

    Brain4Agent

    Memory & context

    Long-term memory for agents: SQLite-backed BM25 store with CJK-first retrieval, CLI, Python library, MCP server and agent skill.

    GH#mcp-server#cli#any-agent1 post · 13 days ago★ 172 +14/wk

  8. #8

    deja

    Memory & context

    Local proxy that dedupes and cache-safely evicts stale context in coding agent sessions, cutting input token bills.

    GH#claude-code#codex#any-agent1 post · 12 days ago★ 15 +1/wk

  9. #9

    Graft

    Memory & context

    Open-source context layer that builds a linked markdown code graph so coding agents skip repo re-exploration, saving tokens and time.

    GHHN#claude-code#codex#cursor6 posts · 19 days ago★ 11.4K +115/wk

  10. #10

    CMI — Context Manager Intelligence

    Memory & context

    Local zero-LLM proxy that folds stale context in coding-agent sessions, cutting tokens sent by ~30% with byte-exact recall.

    GH#claude-code#codex#macos1 post · 12 days ago★ 11

  11. #11

    Graphify

    Memory & context

    CLI that parses code and docs into a queryable knowledge graph, installed as a skill for AI coding assistants.

    GHHN#claude-code#cursor#codex5 posts · Oct 21★ 151.3K +3.9K/wk

  12. #12

    cogit

    Memory & context

    Claude Code plugin that records each agent message as a thought commit and links code commits to the decisions behind them.

    GH#claude-code#plugin#cli1 post · 25 days ago★ 213 +41/wk

  13. #13

    Renor

    Memory & context

    Indexes Markdown notes into SQLite and exposes search, recall and remember tools to coding agents over MCP.

    HN#claude-code#codex#mcp-server3 posts · 17 days ago★ 55 +3/wk

  14. #14

    mnemonic-proxy

    Memory & context

    Transparent HTTP proxy that archives and recalls agent history verbatim, letting coding agents run long sessions within a small model contex

    HN#claude-code#codex#local-models1 post · 27 days ago★ 81 +11/wk

  15. #15

    OptScale Kompressor

    Memory & context

    Rust proxy that deterministically compresses coding-agent prompts to cut token costs while keeping removed content retrievable.

    HN#codex#claude-code#cli1 post · Oct 19★ 65 +18/wk

  16. #16

    Thenwise

    Memory & context

    Claude Code plugin storing past chats, commits and notes in a local SQLite store with search tools and hooks for the agent.

    GH#claude-code#plugin#local-models1 post · 25 days ago★ 36 +4/wk

  17. #17

    funes

    Memory & context

    A single-binary memory layer that indexes coding agent session traces locally and lets agents recall past decisions with provenance, optiona

    HN#claude-code#codex#pi3 posts · Sep 8★ 600 +7/wk

  18. #18

    SuperLocalMemory

    Memory & context

    Local, governed long-term memory for AI agents, shared across Claude Code, Codex, Cursor and other MCP clients.

    HN#claude-code#codex#cursor2 posts · Oct 14★ 290 +14/wk

  19. #19

    agent-memory

    Memory & context

    Long-term memory runtime for AI agents using plain Markdown files as the source of truth with local ranked retrieval and a sleep-time consol

    GH#claude-code#codex#cli2 posts · Oct 18★ 4.3K +7/wk

  20. #20

    backpass

    Memory & context

    Improves agent memory files (AGENTS.md, CLAUDE.md, skills) by mining local coding-agent session transcripts and proposing evidence-gated edi

    HN#claude-code#codex#cli1 post · Sep 14★ 1.5K +25/wk

  21. #21

    hermes-vcc-engine

    Memory & context

    Hermes Agent plugin that compacts conversation context in milliseconds without an LLM call, preserving commands, errors and files.

    GH#plugin#local-models#cli1 post · Oct 14★ 30 +2/wk

  22. #22

    agent-knowledge-mcp

    Memory & context

    MCP server providing agents shared long-term memory as Markdown files, with lexical search, remote access, and multi-device sync.

    GH#mcp-server#any-agent#local-models1 post · Oct 13★ 32 +2/wk

  23. #23

    Aletheia

    Memory & context

    Memory for coding agents where every memory carries code evidence that is re-checked on recall, hiding stale facts from context.

    GH#claude-code#codex#cursor1 post · Oct 21★ 17 +1/wk

  24. #24

    dsh-auto-memory

    Memory & context

    Claude Code-style auto-memory plugin for DeepSeek Harness: typed Markdown memory files with a MEMORY.md index injected into prompts.

    GH#plugin#local-models#cli1 post · Sep 22★ 137 +5/wk

  25. #25

    compress

    Memory & context

    CLI proxy that trims coding-agent tool call output with a fine-tuned Qwen model to reduce input tokens and API costs.

    HN#codex#cli#qwen2 posts · Sep 30★ 65

Filters

License
Category
Works with
Show 7 more
Models
Show 1 more
Platform
Interface
Show 2 more
GitHub stars
Language
First seen
Clear all