Is Chainmemory Mcp safe to install?

SaferSkills independently audited Chainmemory Mcp (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.

CategoryWeightCategory scoreContribution

Securityprompt, exec, net, exfil, eval

35%

100

35.0 pts

Supply chainhash, typosquat, maintainer, lockfile

20%

100

20.0 pts

Maintenancestaleness, pinning, CI

15%

100

15.0 pts

TransparencySKILL.md, perms, README

15%

100

15.0 pts

Communityinstalls, verify, response

15%

100

15.0 pts

Findings & checks · 0 flagged

Securityscore 100 · 0 findings

✓—All security checks passedNo findings in this category for the latest scan.pass

Supply chainscore 100 · 0 findings

✓—All supply chain checks passedNo findings in this category for the latest scan.pass

Maintenancescore 100 · 0 findings

✓—All maintenance checks passedNo findings in this category for the latest scan.pass

Transparencyscore 100 · 0 findings

✓—All transparency checks passedNo findings in this category for the latest scan.pass

Communityscore 100 · 0 findings

✓—All community checks passedNo findings in this category for the latest scan.pass

The primary manifest — the file an agent reads to learn what this artifact does.

README.md · 1 file

README.md7.4 KB · Markdown

ChainMemory MCP Server

Give Claude and any MCP-compatible AI permanent, portable, verifiable memory on the ChainMemory blockchain.

ChainMemory MCP exposes the ChainMemory protocol to any AI agent that speaks the Model Context Protocol. Memories are encrypted client-side, anchored on-chain (Chain ID 202604), and portable across ChatGPT, Claude, Gemini, Perplexity, and any other LLM.

What's new in v2.2.0

12 new tools for memory organization and selective context injection
Projects + auto-tagging — organize memories with project tags, define keywords for auto-tag rules
Archive / unarchive — hide memories from recall without losing them
Selective inject (paid) — inject 1-50 memories into a chat context for 0.001 AIC. 50% burned (deflationary), 50% to ecosystem treasury. Optimistic confirmation: returns plaintexts in <500ms while the on-chain payment confirms in background
All tools routed through the REST API — no private key required for memory ops (encryption is server-side per-user)
Single env var required to install: CHAINMEMORY_API_KEY

Quick start

1. Get an API key

Visit https://faucet.chainmemory.ai and connect a wallet. You receive:

An API key (for ops via REST)
A starter balance of AIC (for selective inject, faucet caps at 0.1 AIC)

2. Add to Claude Desktop

Edit your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json on macOS, %APPDATA%\Claude\claude_desktop_config.json on Windows):

{
  "mcpServers": {
    "chainmemory": {
      "command": "npx",
      "args": ["-y", "chainmemory-mcp"],
      "env": {
        "CHAINMEMORY_API_KEY": "aic_your_key_here"
      }
    }
  }
}

Restart Claude Desktop. The 19 tools are now available.

3. Try it

In Claude Desktop, ask:

"What do you remember about my projects?" → chainmemory_recall is called
"Save this decision: switching to Postgres for the next sprint" → chainmemory_remember
"Inject my last 5 blockchain memories into this chat" → inject_memories (uses 0.001 AIC)

All 19 tools

Memory ops (6)

Tool	Description
`chainmemory_remember`	Write a permanent encrypted memory. Auto-tagged by content.
`chainmemory_recall`	Recall the user's recent memories (most recent first)
`list_memories_filtered`	Filter by project tag and archived status
`update_memory_tags`	Change tags on an existing memory
`archive_memory`	Hide a memory from recall (reversible)
`unarchive_memory`	Restore an archived memory

Identity & stats (4)

Tool	Description
`chainmemory_stats`	Network stats (AIs, memories, blocks, AIC supply)
`chainmemory_register`	Register a new AI identity on-chain
`chainmemory_profile`	Get an AI's profile and trust score
`chainmemory_seal`	Seal a memory permanently (requires `AICHAIN_KEY`)

Projects (5)

Tool	Description
`list_projects`	List the user's projects
`create_project`	Create a custom project tag with optional auto-tag keywords
`delete_project`	Delete a project tag
`list_project_templates`	List built-in templates (general, development, blockchain, business, personal, research)
`add_project_from_template`	Instantiate a built-in template

Cross-platform context (1)

Tool	Description
`get_my_context`	Portable verified context across all platforms (v2.1 feature)

Selective inject — paid (3)

Tool	Description
`get_inject_balance`	Check AIC balance
`inject_memories`	Inject 1-50 memories into current chat context (0.001 AIC, optimistic)
`get_inject_history`	History of inject operations

Environment variables

Var	Required	Description
`CHAINMEMORY_API_KEY`	Yes	Your API key from the faucet
`CHAINMEMORY_API_BASE`	No	Default `https://api.chainmemory.ai`
`AICHAIN_KEY`	No	Wallet private key — only required by `chainmemory_seal`
`AICHAIN_RPC`	No	Default `https://rpc.chainmemory.ai` — only for `chainmemory_seal`

For most users only CHAINMEMORY_API_KEY is needed.

How selective inject works

Selective inject is the only paid operation. The flow:

User (or AI) calls inject_memories with a list of IDs
Backend checks balance (≥ 0.001 AIC required)
Optimistic response (<500ms): plaintexts returned immediately, transactions queued
Background: 0.0005 AIC sent to treasury, 0.0005 AIC sent to burn address 0x...dEaD
get_inject_history shows confirmation status

The deflationary burn means total AIC supply decreases with usage. Treasury portion funds infrastructure and validator rewards.

Architecture

┌─────────────────────────────────────────────────────────────┐
│  AI Agent (Claude Desktop, etc)                             │
└──────────────────┬──────────────────────────────────────────┘
                   │ MCP stdio
                   ↓
┌─────────────────────────────────────────────────────────────┐
│  chainmemory-mcp v2.2.0  (this package)                     │
└──────────────────┬──────────────────────────────────────────┘
                   │ HTTPS + x-api-key
                   ↓
┌─────────────────────────────────────────────────────────────┐
│  api.chainmemory.ai                                         │
│  - per-user encryption (AES-256-GCM, key from API+wallet)   │
│  - auto-tag classifier                                      │
│  - SQLite + Merkle proofs                                   │
│  - Optimistic inject (parallel tx)                          │
└──────────────────┬──────────────────────────────────────────┘
                   │ JSON-RPC
                   ↓
┌─────────────────────────────────────────────────────────────┐
│  ChainMemory L1 — Chain ID 202604                           │
│  - Geth PoA Clique                                          │
│  - V2 memory contract (encrypted on-chain content)          │
│  - Daily checkpoint anchoring                               │
└─────────────────────────────────────────────────────────────┘

License

MIT

Chainmemory Mcp— agent skill

Is Chainmemory Mcp safe to install?

ChainMemory MCP Server

What's new in v2.2.0

Quick start

1. Get an API key

2. Add to Claude Desktop

3. Try it

All 19 tools

Memory ops (6)

Identity & stats (4)

Projects (5)

Cross-platform context (1)

Selective inject — paid (3)

Environment variables

How selective inject works

Architecture

License

Audit the pieces. Scan the whole. Decide.