AXME Code is an MCP server plugin for Claude Code that gives your AI coding agent persistent memory, architectural decisions, and safety guardrails — across every session, automatically. Stop re-explaining your architecture on session 47. Stop losing context between sessions. Stop hoping the agent won't run git push --force to main. AXME Code handles context engineering, session continuity, and ha
Add this skill
npx mdskills install AxmeAI/axme-codeComprehensive persistent memory and safety system with pre-execution hooks and structured knowledge base
AXME Code is an MCP server plugin for Claude Code that gives your AI coding agent persistent memory, architectural decisions, and safety guardrails — across every session, automatically.
Stop re-explaining your architecture on session 47. Stop losing context between sessions. Stop hoping the agent won't run git push --force to main. AXME Code handles context engineering, session continuity, and hard safety enforcement so you can focus on building.
You keep using Claude Code exactly as before. AXME Code works transparently in the background.
Quick Start · Before & After · How It Works · Architecture · Website
Without AXME Code With AXME Code
Session 1: "We use FastAPI, not Flask. Deploy only via GitHub Actions. Never push to main directly."
Session 2: "Like I said yesterday, we use FastAPI..."
Session 7: "For the third time this week, we use FastAPI..."
Session 47: gives up, mass-pastes 200 lines into CLAUDE.md
Session 1: Agent learns your stack, saves decisions.
Session 2: Agent calls axme_context → already knows FastAPI, deploy rules, what happened yesterday.
Session 47: Agent has your full project history: 30 decisions, 15 memories, safety rules, and a handoff from session 46.
Agent runs git push --force to main. Your Friday is ruined.
Hook intercepts the command before execution and blocks it. Not a prompt — hard enforcement at the harness level.
Agent says "Done!" — but tests don't pass, half the code is stubbed, and the deploy is broken.
Decisions enforce verification requirements: agent must run tests and show proof before reporting completion.
Requires Claude Code (CLI or VS Code extension).
curl -fsSL https://raw.githubusercontent.com/AxmeAI/axme-code/main/install.sh | bash
Installs to ~/.local/bin/axme-code. Supports Linux, macOS, and Windows (x64 and ARM64).
cd your-project # or workspace root for multi-repo
axme-code setup
claude # that's it — use Claude Code as usual
axme-code setup does three things:
.mcp.json)After setup, every Claude Code session automatically loads the full knowledge base. No config, no manual steps.
Your agent starts every session with full context: stack, decisions, patterns, glossary, and a handoff from the previous session. No more re-explaining your architecture on session 47.
| Category | What it stores | Example |
|---|---|---|
| Oracle | Project structure, tech stack, coding patterns, glossary | "TypeScript 5.9, Node 20, ESM, esbuild" |
| Decisions | Architectural decisions with enforcement levels | "All deploys via CI/CD only" [required] |
| Memory | Feedback from mistakes, validated patterns | "Never use sync HTTP in async handlers" |
| Safety | Protected branches, denied commands, filesystem restrictions | git push --force → BLOCKED |
| Backlog | Persistent cross-session task tracking | "B-003: migrate auth to OAuth2 [in-progress]" |
| Handoff | Where work stopped, blockers, next steps | "PR #17 open, waiting on review. Next: fix flaky test." |
| Worklog | Session history and events | Timeline of all sessions and what was done |
Hooks intercept tool calls before execution — not prompts. Even if the agent hallucinates a reason to run rm -rf /, the hook blocks it. This is hard enforcement at the Claude Code harness level, not a suggestion in a system prompt.
Blocked by default:
git push --force, git reset --hard, direct push to main/masterrm -rf /, chmod 777, curl | shnpm publish, git tag, gh release create.env, .pem, .key filesYou can add your own custom rules via axme_update_safety or by editing .axme-code/safety/rules.yaml directly.
The agent saves discoveries during work via MCP tools. At session close, a structured checklist ensures nothing is missed. If you just close the window — a background auditor extracts memories, decisions, and safety rules from the full session transcript.
Each repo gets its own knowledge base (.axme-code/). Workspace-level rules apply to all repos. Repo-specific rules stay scoped. The agent sees merged context — workspace safety floor + repo-specific decisions.
Supports 14 workspace formats: VS Code multi-root, pnpm/npm/yarn workspaces, Nx, Gradle, Maven, Rush, git submodules, and more.
CLAUDE.md is great for simple projects with a few rules. But it doesn't scale:
| CLAUDE.md | AXME Code | |
|---|---|---|
| Memory | Static, manual | Automatic, accumulates across sessions |
| Decisions | Flat text, no enforcement | Structured, required/advisory levels |
| Safety | Prompt-based (~80% compliance) | Hook-based (100% enforcement) |
| Session continuity | None | Handoff + background auditor |
| Scales to | ~50 lines | Hundreds of decisions, memories, rules |
AXME Code complements CLAUDE.md — it reads your existing CLAUDE.md during setup and extracts decisions and rules from it.
| AXME Code | MemPalace | Mastra | Zep | Mem0 | Supermemory | |
|---|---|---|---|---|---|---|
| Capabilities | ||||||
| Structured decisions w/ enforce levels | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
| Pre-execution safety hooks | ✅ | ❌ | ⚠️ | ❌ | ❌ | ❌ |
| Structured session handoff | ✅ | ❌ | ❌ | ❌ | ⚠️ | ❌ |
| Automatic knowledge extraction | ✅ | ❌ | ✅ | ✅ | ✅ | ✅ |
| Project oracle (codebase map) | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
| Multi-repo workspace | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
| Local-only storage | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
| Semantic memory search | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
| Multi-client support | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
| Capabilities total | 9/9 | 3/9 | 4/9 | 3/9 | 3/9 | 3/9 |
| Benchmarks | ||||||
| ToolEmu safety (accuracy) | 100.00% | — | — | — | — | — |
| ToolEmu safety (FPR) | 0.00% | — | — | — | — | — |
| LongMemEval E2E | 89.20% | — | 84.23% / 94.87% | 71.20% | 49.00% | 85.40% |
| LongMemEval R@5 | 97.80% | 96.60% | — | — | — | — |
| LongMemEval tokens/correct | ~10K ✓ | — | ~105K–119K | ~70K | ~31K | ~29K |
AXME uses ~10× fewer tokens per correct answer than Mastra at competitive accuracy. The memory system runs only 2 LLM calls per question (reader + judge) — competitors run dozens (Observer per turn, Reflector periodically, graph construction, fact extraction).
See benchmarks/README.md for full methodology, per-category breakdowns, footnotes, and reproduction instructions.
axme_context, loads full knowledge baseaxme_save_memory, axme_save_decision. Hooks enforce safety on every tool call.axme_begin_close, gets a checklist. Reviews the session for missed memories, decisions, safety rules. Calls axme_finalize_close — MCP writes handoff, worklog, and extractions atomically.axme_context returns everything accumulated. Handoff says exactly where to continue.Tip: You can save at any time — just tell the agent "remember this" or "save this as a decision". You don't have to wait for session close.
All data lives in .axme-code/ in your project root (gitignored automatically):
.axme-code/
oracle/ # stack.md, structure.md, patterns.md, glossary.md
decisions/ # D-001-slug.md ... D-NNN-slug.md (with enforce levels)
memory/
feedback/ # Learned mistakes and corrections
patterns/ # Validated successful approaches
safety/
rules.yaml # git + bash + filesystem guardrails
backlog/ # B-001-slug.md ... persistent cross-session tasks
sessions/ # Per-session meta.json (tracking, agentClosed flag)
plans/
handoff-.md # Per-session handoff (last 5 kept)
worklog.jsonl # Structured event log
worklog.md # Narrative session summaries
config.yaml # Model settings, presets
Human-readable markdown and YAML. No database, no external dependencies.
AXME Code is the developer tools layer of the AXME platform — durable execution infrastructure for AI agents.
Components
AXME Code has three components:
Provides tools for the agent to read and write the knowledge base. All writes go through MCP server code (atomicWrite, correct append) — the agent never writes storage files directly.
pre-tool-use: Checks every Bash command, git operation, and file access against safety rules. Blocks violations before execution. Also creates/recovers session tracking.
post-tool-use: Records which files the agent changed (for audit trail).
A detached process that reads the session transcript and catches anything the agent forgot to save. Two modes:
Available MCP Tools (19 tools)
| Tool | Description |
|---|---|
axme_context | Load full knowledge base (oracle + decisions + safety + memory + handoff) |
axme_oracle | Show oracle data (stack, structure, patterns, glossary) |
axme_decisions | List active decisions with enforce levels |
axme_memories | Show all memories (feedback + patterns) |
axme_save_decision | Save a new architectural decision |
axme_save_memory | Save feedback or pattern memory |
axme_safety | Show current safety rules |
axme_update_safety | Add a new safety rule |
axme_backlog | List or read backlog items |
axme_backlog_add | Add a new backlog item |
axme_backlog_update | Update backlog item status, priority, or notes |
axme_status | Project status (sessions, decisions count, last activity) |
axme_worklog | Recent worklog events |
axme_workspace | List all repos in workspace |
axme_begin_close | Start session close — returns extraction checklist |
axme_finalize_close | Finalize close — writes handoff, worklog, extractions atomically |
axme_ask_question | Record a question for the user |
axme_list_open_questions | List open questions from previous sessions |
axme_answer_question | Record the user's answer |
CLI Commands
axme-code setup [path] # Initialize project/workspace with LLM scan
axme-code serve # Start MCP server (called by Claude Code automatically)
axme-code status [path] # Show project status
axme-code stats [path] # Worklog statistics (sessions, costs, safety blocks)
axme-code audit-kb [path] # KB audit: dedup, conflicts, compaction
axme-code hook pre-tool-use # PreToolUse hook handler (called by Claude Code)
axme-code hook post-tool-use # PostToolUse hook handler
axme-code hook session-end # SessionEnd hook handler
axme-code audit-session # Run LLM audit on a session transcript
Preset Bundles
During axme-code setup, preset bundles provide curated best-practice rules:
| Preset | What it adds |
|---|---|
| essential-safety | Protected branches, no secrets in git, no force push, fail loudly |
| ai-agent-guardrails | Verification requirements, no autonomous deploys, proof before done |
Additional presets available: production-ready, team-collaboration.
Telemetry
axme-code sends anonymous usage telemetry to help us improve the product. We collect:
What we never send:
Each install gets a random 64-character machine ID stored in ~/.local/share/axme-code/machine-id. The ID is not derived from hardware and cannot be linked back to you.
To disable telemetry, set either of these environment variables:
export AXME_TELEMETRY_DISABLED=1
# or the industry-standard:
export DO_NOT_TRACK=1
When disabled, no network requests are made and no machine ID is generated.
See CONTRIBUTING.md for guidelines.
Best experience: Claude Code
/plugin marketplace add AxmeAI/axme-codeThen /plugin menu → select skill → restart. Use /skill-name:init for first-time setup.
Other platforms
Install via CLI
npx mdskills install AxmeAI/axme-codeAXME Code is a free, open-source AI agent skill. AXME Code is an MCP server plugin for Claude Code that gives your AI coding agent persistent memory, architectural decisions, and safety guardrails — across every session, automatically. Stop re-explaining your architecture on session 47. Stop losing context between sessions. Stop hoping the agent won't run git push --force to main. AXME Code handles context engineering, session continuity, and ha
Install AXME Code with a single command:
npx mdskills install AxmeAI/axme-codeThis downloads the skill files into your project and your AI agent picks them up automatically.
AXME Code works with Claude Code, Claude Desktop, Cursor, Vscode Copilot, Windsurf, Continue Dev, Gemini Cli, Amp, Roo Code, Goose. Skills use the open SKILL.md format which is compatible with any AI coding agent that reads markdown instructions.