Never lose context to compaction. Archive pre-compaction transcripts, restore working state, and let the model dig back into what it forgot — MCP tools for Claude Code.

An MCP server for context compaction and recycling in Claude Code

A Codex plugin and local MCP server for deliberate context compaction in VS Code, with a verified task handoff and one-time automatic continuation.

Progressive Skill Disclosure & Context Compaction Engine for Autonomous AI Coding Agents

Context compaction and safety gating for AI agents via TypeSafe Jev: keeps messages verbatim, no summarization. OpenAI, Anthropic, LangChain, CLI, MCP.

Local-first MCP server + CLI that gives terminal AI agents an endless semantic memory via Anchored Compacting, keeping the context window small.

Regression-Aware Context Compaction & Witness Replay for Coding Agents (CEGAR-Harness)

Context optimization middleware for LLM agents — a dynamic tool registry, result masking, token budgeting and smart compaction, so a long session stays inside its window instead of failing for length.

Zero context compaction for Claude Code & LLM agents. Persistent memory powered by OS primitives (demand paging, kswapd eviction, mlock pinning). Single SQLite file, MCP-native, multi-agent shared.

Free MCP bridge for agents that lose context: resume after compaction, hand off between runtimes, recover from failure. No API key, no account, no payment.

Local decision ledger for AI coding agents — keeps why a choice was made and which paths are dead ends, and re-injects it after context compaction. Plain Markdown, zero dependencies, works across Claude Code / Codex / Cursor / Copilot / Gemini.

Claude Code context manager + memory engine. Primary install is as a Claude Code plugin (see README). Also usable standalone as a zero-dep CLI: token monitoring against a model-aware quality ceiling, tailored /compact prompt + auto-clipboard, JSONL gzip b

Context engine prioritizing RTFM search results into compact context packs

A token counter overlay for Claude. Warns you before context compaction hits so you can save what matters.

Give agents fast, accurate, and compact context on your Python codebases

⚜️ An MCP server for context compaction and recycling in Claude Code

Context continuity for Claude Code & Grok — structured handoffs keep agent memory intact across compaction.

Geospatial Context is an MCP server that provides agents with reverse geocoding, nearby OpenStreetMap features, and a compact sourced location analysis.

Persistent local memory for AI coding agents. Preserve decisions and project context across sessions, compaction, and tools.

Compact coding-agent context without summarizing it away. Verbatim, offline, with a receipt and recall for every cut. CLI, MCP server, Claude Code plugin. MIT.

⚡ High-performance, compact Linear client & MCP server for AI agents. Reduces prompt context payload by 70–80% with persistent auth, CLI, skills & git hooks.

Comment directives in, compact AI context out — a drift-aware code graph for JS/TS that gives Claude, Codex & other assistants a small query pack instead of the whole tree. CLI + MCP server.

Context optimization middleware for LLM agents — a dynamic tool registry, result masking, token budgeting and smart compaction, so a long session stays inside its window instead of failing for length.

Cross-platform CLI toolkit and local MCP server for AI coding agents. Optimizes context windows and token consumption by turning raw outputs into compact, deterministic Markdown/JSON reports.

Progressive Skill Disclosure & Context Compaction Engine for Autonomous AI Coding Agents. Reduces prompt token bloat by 96.7% using JIT discovery, ephemeral schema injection, and semantic receipts.

Local-first Model Context Protocol (MCP) memory layer for Codex CLI/Desktop, Claude Code, Gemini CLI, Qwen/DeepSeek/Ollama and agent workflows. SQLite + FTS5 compact context packs, token savings, read-only mode, no external memory server.

Session continuity and persistent memory for Claude Code, OpenAI Codex CLI & Google Gemini CLI. Auto-injects past context on start, hands over hot files + build status across compaction, shares one local db across all three. Every feature is opt-in via `p

MCP server for Claude Code, powered by the v4 semantic runtime: builds persistent, budget-aware repository intelligence from AST/graphs, serves bounded context packages, preserves agent state across /compact, and provides end-to-end edit impact analysis in a single Rust binary.

Give AI coding agents a map before they read your code. Statically maps your codebase structure and dependencies, enriches components with LLM summaries, and links Git commits/Jira tickets to nodes. Optimizes agent context windows and token usage with compact, queryable JSON blueprints.

Session continuity for AI coding agents — your agent picks up where it left off. Persistent memory for Claude Code, OpenAI Codex CLI & Google Gemini CLI, sharing one local db: auto context injection, compaction handover, semantic search (KO/EN/JA), error→solution recall. 100% local, zero config, zero API cost.

Local-first, file-native long-term memory hub: preflight, compact context injection, citations, human-approved commits, temporal validity, dedupe/merge, retrospective reflection, write-risk flags, a derived SQLite index, per-workspace scheduled distillati

See what your coding agent sees. A Rust-powered code/ontology/knowledge map shared between you and Claude Code, Codex or Copilot: compact context for the agent, a visual map for the human.

Save your coding agent's context for code: discover and run tasks across Just, Go, Make, CMake, and Docker through one local MCP server, with compact run results and full logs on demand. Put guardrails around your coding agents' commands.

Scan C/C++/Go/Python/Java/Rust/etc. codebases and generate directed call graphs with call ordering, conditions, conditional compilation paths, concurrency analysis, and confidence classification. Designed to give LLM agents global-to-local code understanding — first load a compact context pack, then drill into specific functions on demand.

REST + MCP context-compression layer for Japanese institutional public data. jpcite turns long PDFs, official pages, and search results into compact Evidence Packets with source URLs, fetched timestamps, known gaps, and compatibility/exclusion rules before downstream AI agents draft answers. 3 yen/billable unit metered (3.30 tax-incl), anonymous 3/day per IP free.

Persistent codebase memory for AI coding agents — Claude Code, Codex, Cursor, Cline, Continue and any MCP client. Learns which files go together from how you work and serves compact, ranked context: the gold file on 40 of 40 public-benchmark queries to keyword ripgrep's 32 of 40, at 31-67× fewer tokens than ripgrep and 51-242× fewer than pasting every source file (a larger 150-question eval misses 10). Local-first, no telemetry by default.

93 agentic MCP tools + 49 specialist agents for every major AI CLI — self-learning, no API keys. Veto gives Claude Code, Codex CLI, Gemini CLI, Cursor, Windsurf, Zed, and JetBrains a council of specialist agents, persistent cross-platform memory, a self-learning router, CI/CD gates, and dependency-hallucination + decision-drift guards — all powered by your existing AI subscription. No separate API keys, no extra billing. **Highlights** - **Council debate** — multi-agent GREEN/YELLOW/RED verdicts on risky changes (LLM-backed via MCP Sampling, with deterministic fallback) - **Cross-platform memory & session handoff** — save on Claude, resume on Gemini or Codex; nothing needs re-explaining - **Self-learning router** — improves task routing from recorded outcomes - **Guards & gates** — CI gate, diff review, secrets scan, dependency-hallucination verification, decision-drift enforcement - **Compact mode** — advertise 9 tools instead of 93 and save ~16K context tokens; every tool stays callable **Links**: [GitHub](https://github.com/jigyasudham/veto) · [npm](https://www.npmjs.com/package/@jigyasudham/veto)

SAP ABAP MCP is a local-first Model Context Protocol server that enables AI coding agents to develop, test, analyze, and operate SAP ABAP systems through the official ABAP Development Tools HTTP services. It supports repository search and source editing, structured DDIC workflows, enhancement inspection, paged package navigation, source-backed object creation, semantic services, activation, ABAP Unit and ATC, transports, abapGit, RAP generation, debugging, bounded class and executable-program profiling, runtime dumps, traces and feeds, cross-system comparison, dependency analysis, guarded refactoring, and an opt-in same-origin Screen/GUI Status bridge. Curated compact, development, and assurance presets reduce tool-schema token use without removing capabilities from the default surface. The default local stdio mode can be extended with governed self-hosted Streamable HTTP, API-key roles, OIDC/JWT authentication, request-scoped SAP bearer forwarding, per-session isolation, rate limits, structured audit logs, or a side-effect-free embeddable Node.js API. It also includes browser OAuth Authorization Code with PKCE, a read-only transport assurance CLI and GitHub Action, BTP service-key profile import, and live-SAP evidence tooling. SAP profiles and credentials remain on the user's computer, with passwords and OAuth credentials protected by macOS Keychain or Windows DPAPI. SAP-dependent capabilities are validated against the user's own SAP release, configuration, and authorizations.

Give AI coding agents your Figma design in a compact format: ~94% fewer tokens than raw Figma JSON. Figma plugin + Spring AI MCP server.

Free MCP stdio bridge for agent continuity: resume after compaction, hand off between runtimes, recover from failure. No API key.

[DEPRECATED — migrate to Kapa + Midnight Expert: https://docs.midnight.network/blog/migrating-to-kapa-and-midnight-expert] Model Context Protocol Server for Midnight Blockchain Development

Flatten a Claude Code (or OpenAI Codex CLI) session: bulky tool output the agent already digested moves to a local backup, every word of the conversation stays, and the session resumes lighter — nothing summarized, fully reversible. The lossless alternati

Sovereign Zero-Cloud Node & Human-AI governance engine for Google Antigravity 2.0 & Ubuntu 26.04 LTS: zero-dependency Python 3.14 ASGI web/plugin runtime, 16-pack stack matrix (4 bases + 3 stacks), federated multi-layer mesh, epistemic knowledge corpora & Contextual MCP Server (rcap1 & Docker).

CogmemAi is the safe, persistent memory that keeps Ai in line. It remembers what matters across every session, stops any action that breaks the rules you set before it happens, and reviews what your Ai did against what you asked for. Rule packs: npx cogme

MCP server: locale-aware number formatting (currency, percent, compact).

Token-lean MCP server for Figma: read designs as compact text, render only on demand

An MCP server that lets Claude Code recall past conversations from any project, as clean compact Markdown.

Local CLI and MCP server for compacting test and build output, with retrieval from a redacted local cache.

Your project remembers across coding agents. Compact, cited project memory for Codex, Claude Code, Cursor, and compatible tools.

Async MCP worker for Codex: route bulk code work through cheaper gateways and return compact, verified evidence.

Agent emotional continuity with PAD state, trust, appraisal, and compact emotion logs.

Node.js FHIR MCP server with SMART Backend Services, metadata-aware search/CRUD tools, compact responses, and FHIRPath filtering.

Read-only MCP server giving AI agents the web as compact, ranked, verified evidence. Local cross-encoder reranker, NLI claim verification, semantic cross-source agreement and contradiction detection. No API keys.

Read-only MCP server giving AI agents the web as compact, ranked, verified evidence. Local cross-encoder reranker, NLI claim verification, semantic cross-source agreement and contradiction detection. No API keys.

Lode — keep AI coding on one reviewed product direction across Claude Code, Codex & OpenCode, with compaction-safe recovery and change lineage.

PocketBase CLI for AI agents and people: compact output, full API coverage, and a read-only mode that can be enforced

Local Trello MCP for Claude, Codex and ChatGPT: compact responses, workspace allowlists and read-only opt-in.

Correctness-oriented MCP server and batch CLI for compact structural reads and reviewed TypeScript/JavaScript edits.

Local-first MCP memory server for AI coding agents with compact retrieval and project/global scopes.

Local-first MCP memory server for AI coding agents with compact retrieval and project/global scopes.